Topics in Rational Choice Theory: Altruism, Consequentialism, and Identity

by

Sahar Zahida Akhtar

Department of Philosophy Duke University

Date:_______________________

Approved:

___________________________ Allen Buchanan, co-supervisor

___________________________

Alex Rosenberg, co-supervisor

___________________________ Jesse Prinz

___________________________

Susan Wolf

___________________________ David Wong

Dissertation submitted in partial fulfillment of the requirements for the degree of Doctorate

of Philosophy in the Department of Philosophy in the Graduate School

of Duke University

2008

ABSTRACT

Topics in Rational Choice Theory: Altruism, Consequentialism, and Identity

by

Sahar Zahida Akhtar

Department of Philosophy Duke University

Date:_______________________

Approved:

___________________________ Allen Buchanan, co-supervisor

___________________________

Alex Rosenberg, co-supervisor

___________________________ Jesse Prinz

___________________________

Susan Wolf

___________________________ David Wong

An abstract of a dissertation submitted in partial

fulfillment of the requirements for the degree of Doctorate in Philosophy in the Department of

Philosophy in the Graduate School of Duke University

2008

Copyright by Sahar Zahida Akhtar

2008

iv

Abstract

Rational Choice theory includes a broad body of research that attempts to account

for how people act in a variety of contexts, including economic, political and even moral

situations. By proposing, most generally, that individuals rationally pursue their self-

interests regardless of the context, rational choice has had extensive theoretical and

empirical success, on the one hand, and has also faced wide criticism when applied in a

variety of disciplines, on the other hand. While there is disagreement over what the

defining assumptions of rational choice theory are, in this dissertation I focus on three on

which there is widespread agreement. These three features of rational choice theory are:

its assumption of egoism or self-interest as the central motivation of individuals; its

reliance on consequences as part of a comparative decision-making framework; and

finally, its focus on the individual and not on groups as the methodological and normative

unit of analysis.

In correspondence to these three features, my dissertation is divided into three

parts and explores the separate topics of (I) egoism and altruism; (II) consequentialism

and ethical decision-making; and, (III) individualism and group identity. The dissertation

is not an exercise in showing the extensive problems of rational choice theory, although

there are many. The dissertation rather engages these three topics with differing results,

some of which in fact attempts to revitalize rational choice, or at least features of rational

choice. For the part on altruism, my goal is to demonstrate why the central assumption of

egoism in rational choice theory is problematic. More broadly, I argue for a different way

of defining genuine altruistic motivation. A result of my analysis there is that altruism

v

appears to be more widespread than has been traditionally assumed and is more amenable

to empirical examination. For my discussion on consequentialism, my aim is to re-

characterize rational choice as a mode of moral decision-making. I argue that the moral

agent is one who frequently compares her particular moral ends in a stable fashion and

for this reason cost-benefit analysis is a fully moral framework, one that encourages the

agent to genuinely care for her ends and values. For the topic of individualism and group

identity, my objective is to show how a previously dismissed topic, once unpacked, is

fully consistent with rational choice theory and ought to be of interest to the rational

choice theorist. I show that if the liberal political theorist, including the rational choice

theorist, is to value group identity, the commitment is only limited to valuing a form of

group identity—particularized identity—that is individualist in character.

vi

Contents

Abstract…………………………………………………………………………….……..iv List of Tables……………………………………………………………………….…….ix List of Figures…………………………………………………………………………….x Introduction……………………………………………………………………………….1

Altruism………………………..………………………………………………...3 Consequentialism……………..………………………………………………….8 Identity……………………………………………………………………….....14

Part I: Rational Choice and Altruism 1.Genuine Altruism: What is it and Does it Exist………………………………………22

1.1 Introduction………………………………………………………………...22 1.2 Standard View of Altruism…………………………………………………26 1.3 Evolutionary Arguments……………………………………………………31 1.4 Problems with Evolutionary Arguments for Genuine Altruism……………40 1.5 Conclusion………………………………………………………………….48 1.6 Appendix 1: Are Feelings and Desires Separate?.........................................51

2. Ultimate or Instrumental Altruism? Who Cares…………………………………….. 64

2.1 Introduction………………………………………………………………...64 2.2 Prior Suitability…………………………………………………………….65

2.3 Central Role of Feelings……………………………………………………67

2.3.1 Reason I………………………………………………………….68

2.3.2 Reason II…………………………………………………………73

2.3.3 Reason III………………………………………………………..87

2.4 Conclusion………………………………………………………………….97

vii

2.5 Appendix 2: Applications to Rational Choice……………………………...99

2.5.1 New Avenues for Theoretical Research…………………………99

2.5.2 New Avenues for Empirical Research…………………….……108 Part II: Rational Choice and Consequentialism 3. The Structure of Moral Decision-Making……………………………………….…..115

3.1 Introduction…………………………………………………………….….115

3.2 Against General Value………………………………………………….…118 3.3 Mediating Ends in Our Everyday Lives…………………………………...127

3.4 Considering Each of Our Ends…………………………………………….136

3.5 Mediating Our Moral Values……………………………………………...142

3.6 Conclusion………………………………………………………………...148

4. Cost-Benefit Analysis in Ethical Decision-Making………………………………....151

4.1 Introduction………………………………………………………………..151

4.2 Intrinsic Value and Alienation…………………………………………….152

4.3 Comparing Values in Bioethics……………………………………….......161 4.4 Beyond Alienation: Other Costs and Benefits of Using CBA…………….169 4.5 Conclusion………………………………………………………………...175

Part III: Rational Choice and Individualism 5. Liberal Respect for Identity? Only for Particular Ones…………………………….177

5.1 Introduction……………………………………………………………….177 5.2 Identity and Autonomy…………………………………………………...180

5.3 Including Identity in Respect for Persons……………….………………..186

5.4 The Locality of Respect………………………………….……………….189

5.5 Collective vs. Particularized Identity…………………….……………….197

viii

5.6 Practical Implications of the Debate………………………………….....202

5.7 Conclusion……………………………………………………………….209 Biblography…………………………………………………………….…………….211 Biography………………………………………………………………………….....217

ix

List of Tables Table 1………………………………………………………………………………..51

x

List of Figures Figure 1……………………………………………………………………………….38 Figure 2……………………………………………………………………………….44 Figure 3……………………………………………………………………………….74 Figure 4……………………………………………………………………………….80 Figure 5……………………………………………………………………………….92

1

Introduction

Rational Choice theory includes a broad body of research that attempts to

account for how people act in a variety of contexts, including economic, political and

even moral situations. By proposing, most generally, that individuals rationally

pursue their self-interests regardless of the context, rational choice has had extensive

theoretical and empirical success, on the one hand, and has also faced wide criticism

when applied in a variety of disciplines, on the other hand.1 While there is

disagreement over what the defining assumptions of rational choice theory are, in my

dissertation I focus on three on which there is widespread agreement. These three

features of rational choice theory are: its assumption of egoism or self-interest as the

central motivation of individuals; its reliance on consequences as part of a

comparative decision making framework; and finally, its focus on the individual and

not on groups as the methodological and normative unit of analysis.

In correspondence to these three features, my dissertation is divided into three

parts and explores the separate topics of (I) egoism and altruism; (II)

consequentialism; and, (III) individualism and group identity. The dissertation is not

an exercise in showing the extensive problems of rational choice theory, although

there are many. The dissertation rather engages these three topics with differing

results, some of which in fact attempts to revitalize rational choice, or at least features

1 For general discussions of the assumptions, applications, and potential problems of rational choice theory, see Green and Shapiro, 1994, Pathologies of Rational Choice Theory; a Critique of Applications in Political Science, Yale University Press: New Haven, especially chapter 2 and pp 47-53 and 87-97; and Hausman, Daniel and McPherson, Michael. 1996. Economic Analysis and Moral Philosophy. Cambridge: Cambridge University Press, especially pp.27-29; and Rosenberg, Alexander. 1992. Economics: Mathematical Politics or Science of Diminishing Returns, University of Chicago Press.

2

of rational choice. For the part on altruism, my goal is to demonstrate why the central

assumption of egoism in rational choice theory is problematic. However, for my

discussion on consequentialism, my aim is to re-characterize rational choice as a

mode of moral decision making. For the topic of individualism and group identity,

my objective is to show how a previously dismissed topic, once unpacked, is fully

consistent with rational choice theory and ought to be of interest to the rational choice

theorist.

The emphasis on these topics is distributed according to how extensively each

topic engages the rational choice literature. Thus, I spend the most significant portion

of my dissertation, Chapters 1 and 2, with their subsequent Appendices, discussing

altruism because this issue has been part of a longstanding debate surrounding

rational choice theory—the view that there is no genuine altruism is an essential

feature of the rational choice tradition. I spend relatively less time on the topic of

consequentialism. While consequentialism is an important feature of rational choice,

in Chapters 3 and 4, I demonstrate how a different defense of consequentialism, based

on what it means to really care about ones moral ends, can offer us an alternative

view of the nature of rational choice—not as a systematic descriptive and normative

theory but as a mode of decision making in ethics. Finally, the topic of group identity

is often not coupled with the rational choice literature. However, in Chapter 5, I show

how a unique, and individual-based, understanding of identity allows for its inclusion

in the broad rational choice paradigm.

In addition to different topics, my dissertation also spans three different fields

of philosophy: moral psychology in part I, normative and applied ethics in part II, and

3

political philosophy in the final part. Thus it has the virtue of connecting these

disparate areas of philosophical inquiry into an overall project on rational choice.

Each of the subsequent two topics relates to the first part on altruism in various ways.

For instance, the discussion of consequentialism involves what it means to really care

for one’s particular ends, and has affinities with the discussion of what it means to

really care for others in the context of the debate between egoism and altruism, and

the discussion on identity relates to altruism because identity is often conflated with

altruism. In what follows, I describe each of these topics in greater detail as well as

illuminate some connections between them and explain how these connections as

well as the dissertation more generally will serve as the basis for future work.

Altruism

In Chapters 1 and 2, I discuss the common assumption of egoistic, or self-

interested, motivation in rational choice theory. My analysis does not directly purport

to show that this assumption is incorrect and that altruistic motivation is widespread,

but it does so indirectly. I try to demonstrate that the standard framework for

conceiving of altruism and egoism is misguided. Specifically, I argue that the

standard conception of ‘genuine’ altruistic motivation is insufficient because it omits

any reference to feelings. I then argue for a different way of thinking about genuine

altruistic motivation, one in which feelings have a privileged role. A result of my

analysis is that there seems to be a continuum of motivation, ranging from egoistic to

4

altruistic, and that the altruistic sort appears to be more widespread than has

traditionally been recognized.

Both the rational choice and philosophy literature describe altruistic

motivation in terms of a desire or preference for the improved welfare of another.

Combined with a relevant belief, such as the belief that someone has experienced

harm, a desire or preference for the improvement or protection of another’s welfare

constitutes altruistic motivation. Genuine altruistic motivation (genuine altruism,

henceforth) is furthermore defined in terms of an ultimate desire or preference for the

improved welfare of another, one that cannot be reduced to any desires for the

improvement of one’s own welfare.2 On the other hand, if someone possesses a

desire to help another but the satisfaction of that desire only serves as a means to the

satisfaction of a desire for one’s own welfare this would be a case of reductive, or

instrumental, and not genuine, altruism. If the desire to improve one’s welfare is

ultimate, then this would constitute egoism.3

To see the contrast between egoism and genuine altruism more clearly,

consider an example. If I have a desire to give money to charity, this is an altruistic

motivation. But we have not yet determined whether it is genuine altruism or

reductive altruism. According to the standard view, it would be genuine altruism if

2 This is the view of philosophers who have written extensively on the topic, such as C.D. Broad (1952. Ethics and the History of Philosophy. London: Routledge and Kegan Paul.) , Joel Feinberg (“Psychological Egoism.” In S. Cahn, P. Kitcher and G. Sher (eds.), Reason at Work, pp. 25-35. Sand Diego: Harcourt Brace and Jovanovich), Thomas Nagel (The Possibility of Altruism. Oxford: Oxford University Press), and Elliott Sober and is the view also attributed, although incorrectly, to Joseph Butler (Five Sermons Preached at the Rolls Chapel and A Dissertation Upon the Nature of Virtue, ed., Stephen Darwall, Indianapolis: Hackett Publishing Company, Inc., 1983) 3 Even some rational choice theorists seem to make this kind of a distinction. See for instance Andreoni, J. (1990): Impure Altruism and Donations to Public Goods: A Theory of Warm-Glow Giving. Economic Journal 100: 464-77

5

my desire to give money is an ultimate one—if I have no other desire to give money

to charity other than that it would help others. On the other hand, if I desire to give

money to charity only because I ultimately desire the feeling that I have made a

difference in the world and giving money is a way to make a difference in the world,

then my desire to give money would be a case of reductive altruism, which according

to the standard view, would still constitute egoism. Egoism can consist either in the

pursuit of any kind of external good, such as material goods only or material goods

combined with public approval, distinction or fame, or can involve the pursuit of

pleasure and other internal rewards—the latter being hedonism, one specific form of

egoism.4

While the question of whether there is any altruistic motivation is

uncontroversial—it seems obvious that people do have desires to help others or, more

generally, desires for the improved welfare of others—the answer to the question of

whether there is genuine altruistic motivation is far less obvious. That is, at least if we

maintain the standard definition of genuine altruistic motivation. This is because a

persistent problem for the existence of genuine altruism has been that critics can

always seemingly reduce an altruistic motivation to seemingly egoistic motivations—

to prudential motivational explanations such as “she’s kind to them because otherwise

she would not have relationships with them,” or, especially, to hedonistic

motivational explanations such as “he gives to charity because it feels good to give to

charity,” and “he feeds his children because it hurts to see one’s children go hungry”.

4 Brennan and Pettit argue for instance that the desire for esteem and status play a large role in both economic and political life. (2004) The Economy of Esteem: An Essay on Civil and Political Society. Oxford: Oxford University Press.

6

Arguments for the existence of egoism consisting of external rewards can be easily

refuted by citing examples of people taking action where there are no material or

recognition returns to be had (a prime example is voting in elections). Therefore

hedonism provides the strongest position for the thesis of egoism.

Since internal rewards are opaque to researchers and since egoism is a robust

descriptive thesis about many of our motivations, proponents of reductive altruism

argue that there simply is no need to posit genuine altruism. Researchers who take a

neutral position on the question have noted, as Daniel Dennett does, that “genuine, or

pure, altruism is an elusive concept, an ideal that always seems to evaporate just when

you get in position to reach out to grab it. It isn’t clear what would count as genuine

altruism, and paradox hovers constantly nearby.”5

A variety of empirical, philosophical and evolutionary arguments have been

advanced to demonstrate the existence of genuine altruism. However, in the first two

chapters of my dissertation, I attempt to show why all of these arguments are

unsuccessful. In Chapter 1, I demonstrate that the most plausible evolutionary

argument, put forth by Sober and Wilson, fails. Because they provide a novel

argument for the widespread existence of genuine altruism and because an

evolutionary framework shares important features with rational choice theory, I spend

some time on Sober and Wilson’s argument. I then demonstrate the failure of both the

philosophical arguments and the empirical work that attempt to show that genuine

altruism exists.

5 Dennett, Daniel. 2003. Freedom Evolves, New York, NY: Viking Publishing, p. 194.

7

In each of these lines of research, I argue in the second chapter, the source of

the failure to establish the existence of genuine altruism is the way in which genuine

altruism is defined. Specifically, I argue that while we often think of altruistic

motivation in terms of feelings such as sympathy and empathy, the standard

definition of genuine altruism makes no essential reference to these feelings. Rather,

the standard account makes the presence of an ultimate altruistic desire or preference

the essential feature of genuine altruism. I argue that, for a variety of reasons,

whether or not a desire or preference is ultimate does not satisfactorily address

whether the motivation is genuine. I then argue that the presence of certain kinds of

feelings better indicates genuineness in motivation. Furthermore, I argue that feelings

and desires/preferences are distinct and that feelings can be motivating. My goal is

not to specify the range of feelings that constitute genuine altruism, but rather to show

that certain types of feelings are essential to our conception of genuine altruism.

My analysis of the central importance of feelings to genuine altruistic

motivation has broad implications for rational choice theory and its application to

explain behavior. First, my proposed account of genuine altruism clears away the

conceptual baggage that has notoriously prevented both an acceptance that genuine

altruism exists and efforts to explore the reach and contexts of altruistic motivation.

Second, once we account for the prominent role of feelings in understanding genuine

altruism, and at the same time recognize the limited, and perhaps absent, role of

ultimate desires, several avenues for empirical research are opened up. Finally,

whereas the standard view of altruism does not always seem to illuminate some forms

of behavior, a view that privileges feelings may offer insight. I argue that feelings are

8

separate from desires and preferences and cannot be reduced to them. At the same

time, I maintain that feelings can be motivating but that they do not aim to be realized

or satisfied in the way that desires do. Because they do not have the aim to be

satisfied, they allow us to make better sense of behavior that is expressive rather than

consequential or instrumental—expressive behavior just is behavior that does not

always seek to achieve any outcome, at least at any given point or in any given

circumstance. Altruistic feelings, by causing expressive behavior, therefore seem to

have many applications to domains of social, ethical and political life involving

community, identity, and social movements.

Consequential ism

In Chapter 3 and 4, I focus on another feature of rational choice theory: its

emphasis on consequences as the locus for determining whether a particular action,

policy or institution is good or bad. There are two ways in which consequences

feature in rational choice theory. First, consequences, and in particular wellbeing

measured by some form of utility or preference-satisfaction, explicitly figure in

welfare economics. Welfare economics is the branch of rational choice theory

dedicated to understanding what economic policies or actions make people better off.

Importantly, whether or not someone is better off is determined by the relevant

consequences. Second, and more subtly, consequences figure in the approach of

rational choice. The rational choice theorist believes that to ask whether an action or

policy is a good one is unhelpful unless we know what the alternative to that action or

policy is. Rational choice therefore is the process of choosing the best, however best

9

is to be defined, selection among available options. Furthermore, the rational choice

theorist believes that to ask whether one action or policy is better than another is

already to imply a common measurement from which comparisons can be made. By

being comparative in analysis, rational choice then places the emphasis on

consequences, and specifically, on general values. This is because, as I will

demonstrate in the third chapter, general values can more easily serve as the basis

from which to make comparisons between competing options than can particular

ends.

When it comes to value theory—to our moral options—many moral theorists

ranging from Kantians to virtue ethicists to particularists, deny that consequences

ought to serve as the only or main determinant of our values, or of whether an action

ought to be taken or a disposition nurtured. In Chapters 3 and 4, rather than giving a

defense of consequentialism based on its purported rationality or coherence, I try to

provide a defense of consequentialism based on what it means to really care for one’s

ends.

There is a widespread belief among philosophers that a genuine care or

concern is one that does not reduce to other concerns. A genuine care or concern is

one that has intrinsic, and not merely instrumental, value. Indeed, as discussed above,

it is this belief that has shaped the definition of genuine altruism, of sharply

distinguishing altruism from any form of egoism. Philosophers have argued that

serious psychological and ethical tensions arise when we are ultimately concerned

with values of a general or abstract nature, rather than with our particular ends,

relationships, and projects. For instance, Kantianism and Utilitarianism, with their

10

groundings in duty and maximization of happiness, respectively, are criticized for

being too impersonal, of alienating us from our own personal relationships and plans,

and of leading to a loss of integrity, that is, of the value of being specially responsible

for and acting on one’s deeply held dispositions and motives.6 Most recently,

Michael Smith argues that the morally good person is one who is motivated directly,

and non-instrumentally, by her particular moral ends.7 Just as good lovers must have

a direct concern for their loved ones, good people, Smith argues, must have a direct

concern for equality, justice, honesty, the welfare of people, and other things that they

believe are right.

In Chapter 3, I argue that many of us implicitly try to make decisions about

our ends and relationships in the very way which is charged with being problematic.

Furthermore, I argue that trying to mediate conflicts between one’s particular moral

ends, such as when one is faced with helping the distant poor or protecting the rights

of immigrants, by appealing to more general values, such as the promotion of

wellbeing, is on reflection how we tend to think about the morally good person.

Appealing to general values that are separate from particular ones can encourage

broad reflection about all of one’s particular ends, including how they fit together,

and can reveal their shared importance.

6 For discussions of these and related criticisms, see: Philippa Foot, in “Utilitarianism and the Virtues,” Mind 94 (1985), 196-209; Peter Railton, in “Alienation, Consequentialism, and the Demands of Morality,” Philosophy and Public Affairs, Vol.13, No.2 (Spring 1984); Michael Stocker, in “The Schizophrenia of Modern Ethical Theories,” Journal of Philosophy, 73 (1976), 453-66; and, Bernard Williams, in Moral Luck: Philosophical Papers, 1973-1980. Cambridge: Cambridge University Press, 1981, especially “Persons, character, and morality.” 7 Smith, Michael. 1994. The Moral Problem. Massachusetts: Blackwell Publishing.

11

While relying on general values to sort through conflicts does not necessarily

entail that these more general values have specific consequentialist content, as

described above it does reflect the more subtle reliance on consequentialism as a

mode or approach of ethical thinking; the mode is a comparative one. Furthermore, I

argue with the rational choice theorist that even the question of whether general

values can help us settle conflicts between our particular values and ends in a morally

satisfying way is itself a comparative exercise. The question is not whether an appeal

to general values fully solves the moral problem of conflict in one’s ends, but whether

there are good reasons to think that appealing to general values does so better than

alternative approaches.

In Chapter 4, I extend my discussion of consequentialism to defend the use of

cost-benefit analysis in ethics. Cost-benefit analysis does not always assume that the

ends or values to be compared are consequences, but does possess the mode of ethical

thinking I described above. I argue that while traditional cost-benefit analysis that

relies on some notion of efficiency, such as pareto efficiency, is subject to a wide

range of moral objections, cost-benefit analysis as conceived of a framework in which

to compare our moral ends encourages broad reflection and the consideration of each

alternative value. Furthermore, I argue that far from calling for the maximization of

any value, cost-benefit analysis simply calls for the minimization of costs, or more

minimally, the assessment of costs—and costs are not strictly monetary ones but are

rather the values and ends which are not favored whenever an action is undertaken for

the sake of some other end or value. The context for my discussion on the use of cost-

12

benefit analysis in ethics is bioethics, where claims about intrinsic values, and hence

the seeming incomparability of values, abound.

Critics of both consequentialist theories and cost-benefit analysis charge that

in such theories the value of one’s particular ends reduces to the value of other, more

general ends and this implies that consequentialists do not have the appropriate

attitude or commitment to their particular ends. The appropriate commitment entails

treating one’s ends as intrinsic or ultimate values. My response in these chapters thus

shares a general kind of argument with my response in Chapters 1 and 2 that has

relevance for rational choice and ethics, and it is the following: whether a motivation,

in the context of the debate between egoism and altruism, or a value, in the context of

ethical decision making, is an ultimate one does not always in itself tell us very much

about the genuineness of that motivation or commitment to that value, respectively.

Whereas in the first two chapters I argue that the description of altruism includes

references to feelings, in these chapters the focus is not on feelings but rather on the

comparative mode of ethical thinking. In each case, however, this means that

whether a care or concern is ultimate is not necessarily the critical question to pose.

In future work, I will examine a more explicit relationship between my

analysis in the first and second parts of my dissertation by exploring what is required

of us by treating altruism as value, and not only as a description of our motivations.

Although altruism has received extensive attention in many fields, including

economics, anthropology, psychology, and evolutionary biology, it has received

comparatively less in ethical discourse, and this is especially true when we consider

altruism as a normative value. The differences between normative and descriptive

13

altruism are many, but to begin with I offer a crude, yet important, one: while it is

perfectly consistent with descriptive altruism to love and care for one’s own children

and dismiss everyone else in the world, this is not consistent with normative altruism,

or its most closely related virtue—compassion.

Compassion has been given little attention in Western secular ethics, relative

to Eastern philosophy. Both Adam Smith and David Hume of course recognized our

tendencies for fellow-feeling and for perceiving the suffering of others through our

ability to sympathize, but it is often argued that their notion of sympathy is much

more akin to the way we use the term empathy today, as connoting a capacity for

experiencing what others experience and identifying with others—and whether

empathy then causes a caring or concern for others is unclear. However we parse out

the concepts of empathy vs. sympathy, it can plausibly be argued that neither Smith

nor Hume made caring for others central to their normative theories. Other, more

contemporary Western ethical theories, including Utilitarianism, demand that one do

what is right or virtuous, but there is little emphasis on the imperative to be caring or

compassionate, or, even the most closely related virtue, benevolent. Benevolence can

be defined, according to Philippa Foot, as having “the proper end… the good of

others”.8 It might seem that there is a natural fit for benevolence in contemporary

virtue or care ethics. Unsurprisingly, something like benevolence is central to care

ethics in particular. But the most notable care ethics view, that of Carol Gilligan, is

too limited in terms of its ethical requirements on us.9 Gilligan’s view emphasizes

8 Foot, Philippa. 1985. “Utilitarianism and the Virtues”, Mind 94, 196-209. 9 Gilligan, Carol. 1982. In A Different Voice: Psychological Theory and Women’s Development. Portland, OR: Book News, Inc.

14

the uniqueness of relationships, with little to no ethical space for caring for others

with whom we do not enter relationships. Outside of care ethics, while benevolence

is considered to be an important virtue or duty under many views, it is generally one

among many.

Rather than advocating a particular ethical view that locates compassion at its

center, in future work I will argue that compassion is central to morality in general

and that my proposed way of describing altruism and consequentialism is consistent

with the requirements of making compassionate decisions, and more generally, of

being compassionate. In particular, I will argue that a morality based on compassion

is not one that emphasizes the intrinsic or ultimate value either in a variety of separate

concerns or the virtues, as virtue ethics can be said to do, or in actions or the good

will, as Kantianism does. Rather, compassion-based morality must be broadly

consequentialist, and even more, welfarist, and must involve frequent balancing,

comparing and weighing of values because of the nature of the world in which we

live.

Identity

The final part of my dissertation, Chapter 5, turns to political philosophy to

address an apparent tension between rational choice theory, with its close theoretical

ties with the liberal tradition, and the assignment of special status to social or group

identities, such as religion, culture, and race. The liberal conception of respect for

persons, based on the Kantian idea, has been criticized by communitarian authors

15

such as Michael Sandel and Charles Taylor because, they argue, it fails to include a

rich and accurate account of persons. These critics argue that it is not only the

universal features of persons that are deserving of respect. Our distinct social

identities are also very important—identities that are largely beyond one’s self-

reflection and that are highly interdependent on others. Because of the importance of

identity, communitarians argue that liberal democratic societies have an obligation to

respect the identities of different minority and cultural groups and the failure to do so

amounts to a denial of equal respect for persons who are members of those groups.

Bhikhu Parekh puts this claim succinctly when he writes ‘the liberal is in theory

committed to equal respect for persons. Since human beings are culturally embedded,

respect for them entails respect for their cultures and ways of life.”10

The overwhelming focus in the debate on identity has been on membership,

whether formally or merely associative, in identity groups. The question of whether

groups that are formed around shared characteristics ought to be able to collectively

pursue their interests or collectively exercise their rights, has been not been an

especially contentious matter in liberal discourse.11 Much more contentious is

whether individuals should be accorded special rights or treatment as members of

identity-groups, rather than merely as persons or citizens, or whether an individual’s

10 This is also roughly the view of other prominent political theorists. For instance, writing in regard to culture, Kymlika argues that the liberal principle of equal respect for persons requires the recognition of the rights of certain cultures since culture is considered “a constitutive part of who the person is” (Kymlicka, p. 175) and Taylor argues likewise that an important condition of respecting someone is recognizing that person as being constituted by her culture. (Taylor, Charles. 1992. Multiculturalism and the Politics of Recognition. Princeton: Princeton University Press) 11 As Appiah notes, these collective rights “tend to have more friends…Most people think that it is just fine that Utah or the city of Cambridge or the Catholic church can exercise rights, through the ballot box or (in the case of churches) through whatever consensual internal mechanisms they agree upon.” Appiah, Anthony (2005) The Ethics of Identity. Princeton: Princeton University Press, p. 72-3

16

status as a member of an identity-group should ever trump an individual’s rights,

status, or obligations as a person.12 Because identity is almost always viewed in

terms of group-membership, valuing identity seems to be at odds with the principles

of liberalism and rational choice theory where persons are the ultimate source of

value. In particular, the apparent tension between identity recognition and rational

choice emerges from the individualism that is prominent in rational choice theory.

Beginning with Max Weber, proceeding with Friedrich Hayek, and later

reinforced in the work of game theory, rational choice theorists have asserted the

primacy of individual-level over collective-level explanations of phenomena.13

Methodological individualism is the view that all actions, even those that result in the

collective decisions of societies, associations, businesses, and families, must be

explained by reference to the intentional states of individual people. It is important

not to confuse this doctrine with the kind of social atomism often associated with

Hobbes and which can be described as the view that all of human psychology can be

explained pre-socially or without reference to interaction or engagement with other

individuals. Methodological individualism does not make any assumptions about the

12 It is true that for Kymlicka in particular, the reason we must respect identity-groups such as cultures is because membership in identity-groups provides a context of choice for the individual—meaningful options for a life-plan and a background against which to shape one’s values, norms, and desires. Thus, cultures for instance are important ultimately because individual choice and autonomy matter. However, under Kymlicka’s view individuals would still be granted special rights as members of groups. Members of transitional or waning cultures, he argues, are disadvantaged as compared to the majority cultures with respect to a stable context of choice, and so must be accorded special rights. (Kymlicka, Liberalism, Community and Culture.)

13 See Weber, Max. 1968. Economy and Society, ed. Guenther Roth and Claus Wittich, Berkeley: University of California Press; Hayek, Friedrich von. 1942. “Scientism and the Study of Man I,” Economica, 9: 267-91; Olson, Mancur. 1965. The Logic of Collective Action. Cambridge, MA: Harvard University Press.

17

content of the intentional states of individuals and is fully compatible with the claim

that individual psychology is formed through engagement with others. Since rational

choice theory only depends on methodological individualism and not on atomism, it

therefore remains open to emphasizing the importance of social identity on how

individuals construe themselves and how they conceive the world—that is, as long as

we do not describe that social identity as being irreducibly collective in nature.

Methodological individualism has become closely aligned with the stronger

thesis of normative individualism, or what Pettit calls ‘personalism’, that is prominent

in the tradition of liberalism.14 Normative individualism, as the name implies, is a

doctrine about what constitutes the good. It maintains that whatever is of value about

nations, cultures, races, families or any other group, is something that is valuable for

the individuals they affect. In short, the thesis denies that there are irreducibly

collective goods.15 Although rational choice theory is not always explicitly

committed to normative individualism, there is a wide and significant area of rational

choice—welfare economics—that is at least implicitly committed to the liberal thesis

of normative individualism. Welfare economics is the branch of economics that is

concerned with evaluating different institutions, actions, and policies, on the basis of

normative judgments. The single most important analytical tool in this area has been

the condition of pareto efficiency, which maintains that a state of the world, X, is 14 See Pettit’s discussion for a fuller explanation of these points, pp. 22-30, in Goodin, R. E. and Pettit, Philip (eds.) 1995. A Companion to Contemporary Political Philosophy. United Kingdom: Blackwell Publishers. 15 Goodin argues that the claim that some goods (for instance, cultures, norms and language) are only possible in a society—a non controversial claim—does not show us that once these goods exist their value is not the value for individual persons but rather emergences from some new property of the group or collective itself. Goodin, R. E. 1990. ‘Irreducibly social goods—Comment I’ in Rationality, Individualism and Public Policy, ed. G. Brennan and C. Walsh. Canberra: Australian National University.

18

morally better than another state of the world, Y, if in X at least one person is better-

off and no one is worse-off than in Y. The pareto condition is clearly a fundamentally

individualist account of the good because “the moral desirability of alternative states

of the world is exhausted by the well-offness of persons: any moral goodness is moral

goodness for someone.” 16 Thus, a large part of rational choice theory would deny

that there is value in social identity if identity is described as being a property of a

group or collective itself.

My arguments in the final chapter, however, demonstrate that there is an

important way in which we can value social identity that is compatible with both

forms of individualism. I begin Chapter 5 by arguing that any liberal account of

valuing social identity must be constrained by the liberal understanding of autonomy,

which is the broadly Kantian and Rawlsian view that individuals have a capacity to

value and to determine their conceptions of the good. I argue that if liberalism is to

value identity, it would be subject to ensuring that autonomy is not compromised.

Even though placing importance on autonomy is not necessarily a principle of

rational choice theory, I will show how satisfying the condition of autonomy will only

support valuing a form of social identity that is individualist. In particular, I show that

we can accord value and respect to individuals’ identities without ascribing any

special status to groups. Thus, even though the explicit starting point of the chapter is

liberalism, the result is to countenance a form of social identity that is also fully

consistent with both the methodological and normative individualism present in

rational choice theory.

16 Brennan, Geoff, pp. 125-6, in Goodin and Pettit, 1995.

19

The topic of social identity is related to the earlier chapters on altruism

because, as mentioned before, identity is often conflated with altruism. As I discuss

in my chapters on altruism, altruism refers to the motivation to care for others. Since

empathy is a feeling often associated with caring for others and empathy is also

closely related to the notion of identifying with others, it might seem that identity is a

kind of altruism. However, clear distinctions between altruism and identity emerge in

my dissertation. In particular, that I identify with someone else does not yet tell us

whether I care about, or am motivated by, that person’s wellbeing. In addition, it

seems possible to care about someone without identifying with them in any way, such

as when one cares about other animals or even the environment. Furthermore, as we

see in Chapter 5, when commentators discuss identity they are often referring to the

idea that culture, ethnicity, or race in some important sense constitutes our sense of

who we are. These things, they argue, are a significant part of one’s self-conception.

However, it is not even clear whether it is specific others that one identifies with or

whether it is the norms, values, traditions, and languages particular to a way of life. It

seems that at least some of the time it is the latter that is relevant in identity

discussions and in these cases, it seems identity has nothing to do with altruism.

Conversely, it seems plausible that some times when people express concerns over

identity it is really a case of group altruism. For instance, if someone says that all

Pakistanis are stupid, I may feel hurt. But it is not clear whether I am experiencing

hurt because I am a Pakistani and the negative comments are an attack on who I am,

or whether my feeling hurt is for others I care about who happen to be Pakistani. Of

course it is possible that it is a little of both, but at least some of the claims made on

20

behalf of identity may in fact only reflect that we care about others who happen to be

like us in some relevant sense. In general, I think part of the reason for the conflation

of identity with altruism is because commentators fail to make another distinction—

that between individual and group identity. Once we do so, as I do in Chapter 5, we

find that at least in certain contexts it is easier to distinguish between identity and

altruism. In future work, I will go on to explore the relationship between identity and

altruism in the context of caring about distant others with whom one does not identify

in any meaningful sense.

In addition to this future work, I will also continue to engage in the conceptual

work of distinguishing between not only identity and altruism, but also between

identity and ultimate values, as per my discussion in the second part of my

dissertation. Writers such as Sandel and Williams speak of the deep commitments

that one has as being those things which constitute one’s identity. They suggest that

this relationship between identity and deep commitments accounts for why it is

alienating to treat one’s commitments as instrumental—that is, non-ultimate—values.

But there are some important distinctions between what one values and what

constitutes one’s identity. Art and a deep passion for traveling can be things that we

value very much without these things being part of anyone’s identity. At the same

time, we can have features of our identity that are trivial or even shameful—while we

can consistently and perpetually be ashamed of something that is a part of us, it seems

the same does not hold for something we value; we would come to no longer value it.

Thus, while it may be important to distinguish between what Sandel calls mere

preferences and deep commitments, it also seems important to keep distinct the idea

21

that something is a deep commitment from the idea that something is a constituent of

who a person is.17 While it seems plausible that these are sometimes one and the

same, this does not always seem to be true. Working through these distinctions will

be the subject of future work.

17 Sandel, Michael J. (1982) Liberalism and the Limits of Justice. Cambridge: Cambridge University Press, 2nd edition 1999

22

Part I Rational Choice and Altruism

1 Genuine Altruism: What is it, and Does it Exist?

1. 1 Introduction

Critics of rational choice theory have frequently argued that its assumptions

about self-interest, or egoism, are flawed. These critics claim that we at least

sometimes act to benefit others. Contemporary writers on altruism in a variety of

empirical fields, such as economics, experimental game-theory, and biology, define

altruism in terms of behavior. Philosophical and psychological discourse on the topic

of altruism, however, has been largely concerned with motivational altruism, and in

particular, with genuine motivational altruism.1 The standard way of defining

genuine altruism is to contrast it with reductive altruism: genuine motivational

altruism consists in having an ultimate desire or preference for the welfare of

another—a desire or preference that cannot be reduced to other desires that one has

for oneself. 2 In particular, there has been an emphasis on distinguishing genuine

altruism from hedonism, which is a form of egoism. Hedonism is the view that all of

our motivations ultimately reduce to the motivation to avoid pain and seek pleasure.

1 Even some rational choice theorists seem to make this kind of a distinction. See for instance Andreoni, J. (1990): Impure Altruism and Donations to Public Goods: A Theory of Warm-Glow Giving. Economic Journal 100: 464-77 2 This is the view of philosophers who have written extensively on the topic, such as C.D. Broad (1952. Ethics and the History of Philosophy. London: Routledge and Kegan Paul.) , Joel Feinberg (“Psychological Egoism.” In S. Cahn, P. Kitcher and G. Sher (eds.), Reason at Work, pp. 25-35. San Diego: Harcourt Brace and Jovanovich), Thomas Nagel (The Possibility of Altruism. Oxford: Oxford University Press), and Elliott Sober and is the view also attributed, although incorrectly, to Joseph Butler (Five Sermons Preached at the Rolls Chapel and A Dissertation Upon the Nature of Virtue, ed., Stephen Darwall, Indianapolis: Hackett Publishing Company, Inc., 1983)

23

Because hedonism seems to maintain that not wanting to feel bad when one observes

another suffering is a consistently hedonistic, and therefore egoistic, motivation for

action, genuine altruism has been defined in a way that omits any essential reference

to feelings.

Prima facie, the standard philosophical definition of genuine altruism makes

sense, for it seems we would want to reserve the ascription of altruism only for

motivations that are wholly distinct from egoism. However, in this part of my

dissertation I will argue that the customary way of defining altruistic motivations is

unnecessarily demanding, and more importantly, misguided. The arguments I offer

support a different description of genuine altruism: one in which feelings play a

central role. In particular, I will argue that the emphasis on separating altruism from

hedonism misses the importance of feelings to any conception of altruism. Once we

include feelings in our description of altruism, the strong conceptual distinction that is

said to exist between genuine altruism and hedonism collapses. Rather than

identifying the specific feelings that are involved with altruism, my objective is to

argue that feelings for others are essential to our understanding of genuineness when

it comes to altruistic motivation. In Chapter 2, I will argue that these feelings cannot

be adequately represented by, or cannot be reduced to, preferences and desires for a

variety of reasons, but that they are nonetheless a kind of motivation. I will discuss

there how my analysis of the central importance of feelings to altruistic motivation

has broad implications for rational choice theory and its application to many forms of

social and political behavior.

24

When discussing the possibility of altruism in the context of rational choice

theory, a natural place to start is with evolutionary arguments. Rational choice theory

equates what is rational with what promotes one’s welfare, and welfare is typically

defined in terms of the attainment of material resources, such as money, food, shelter,

and other external goods. Several authors have attempted to sever the connection

between rationality and welfare by arguing that it is sometimes rational to be

altruistic, and in particular, that it is evolutionarily or materially advantageous to be

altruistic. For this reason, across the disciplines many discussions of egoism and

altruism take place within the framework of evolutionary arguments, whether in terms

of game-theory, social-psychology, or evolutionary biology: demonstrating the

plausibility of the emergence or evolution of altruism in natural settings would serve

a long way to show that altruism can be beneficial to the agent in some broad sense.

Perhaps the most notable, systematic, and recent example of this effort is the

argument advanced by Sober and Wilson in their important book Unto Others: the

Evolution and Psychology of Unselfish Behavior. 3

A persistent problem for the existence of genuine altruism as defined by the

standard view has been that critics can always seemingly reduce an altruistic

motivation to so-called egoistic motivations— to prudential motivational explanations

such as “she’s kind to them because otherwise she would not have relationships with

them” or to hedonistic motivational explanations such as “he gives to charity because

it feels good to give to charity” and “he feeds his children because it hurts to see

one’s children go hungry”. Since motivations are opaque to researchers and since

3 Sober, Eliot and Wilson, David Sloan. 1998. Unto Others: The Evolution and Psychology of Unselfish Behavior, Cambridge, MA: Harvard University Press.

25

egoism is a robust descriptive thesis about many of our biologically selected

motivations, proponents of egoism argue that there simply is no need to posit genuine

altruism. Researchers who take a neutral position on the question have noted, as

Daniel Dennett does, that “genuine, or pure, altruism is an elusive concept, an ideal

that always seems to evaporate just when you get in position to reach out to grab it. It

isn’t clear what would count as genuine altruism, and paradox hovers constantly

nearby.”4 In the second half of their book, however, Sober and Wilson (S&W)

maintain that it is plausible that not only altruistic motivation but also genuine

altruistic motivation has evolved in cognitively sophisticated animals such as

humans.5

My arguments concerning the nature of altruism proceed as a response to the

S&W book for several reasons, although I discuss where other works on altruism

overlap with theirs. First, S&W adopt the standard view of the nature of genuine

altruism, the view held by most philosophers, but attempt to give a novel argument to

show that altruism in this sense exists. Second, their argument is an evolutionary one

that is informed by the literature in game-theory, anthropology, and psychology, and

has much in common with rational choice theory. I will argue, however, that their

argument has a major shortcoming that points to a more general problem that

evolutionary arguments for altruism confront when they adopt the standard view of

altruism. Finally, S&W’s discussion spends more time on distinguishing instrumental

from ultimate desires—an important feature of the standard view—than do other

4 Dennett, Daniel. 2003. Freedom Evolves, New York, NY: Viking Publishing, p. 194. 5 Sober, Eliot and Wilson, David Sloan. 1998. Unto Others: The Evolution and Psychology of Unselfish Behavior, Cambridge, MA: Harvard University Press.

26

accounts. The general goal of my arguments, however, is to discredit the egoism

thesis central to rational choice theory. By showing how the standard philosophical

view of altruism—even when it takes great pain to distinguish between genuine and

reductive altruism and shares features in common with rational choice theory—

misses something important about the nature of altruism, I hope to demonstrate that

genuine altruism is not an elusive and singular phenomenon but a potentially

significant range of motivations behind many of our actions. In order to do this, I

must not only offer at least preliminary arguments towards a particular understanding

of altruism, but must also explain why the standard view, by putting the emphasis on

the wrong distinction, stacks the deck in favor of the egoism thesis of rational choice.

1 .2 Standard View of Genuine Altruism

Under the standard view, and in line with common belief/desire folk

psychology, what triggers or causes an altruistic preference is a belief, such as the

belief that someone has been harmed.6 The combination of belief and desire

constitutes a complete motivation to act. As will be discussed at greater length below,

philosophers argue that the motivational phenomenon of altruism involves a desire.

Furthermore, and important to this discussion, in order for a case of altruistic

motivation to be considered ‘genuine’, it must consist in an ultimate desire. Put in its

simplest formulation, an ultimate desire represents something that we want for its

own sake, whereas something desired instrumentally is desired because it is a means 6 Rosenberg, Alexander. 1992. Economics: Mathematical Politics or Science of Diminishing Returns, University of Chicago Press; 1998. “Folk Psychology” in J. B. Davis, ed., Handbook of Economic Methodology, Aldershot, Elgar; 1996. "Philosophy of Economics" in Audi, R., Cambridge Dictionary of Philosophy, Cambridge University Press, pp. 582-583; Also see Notes 4 and 5

27

to satisfying another desire, perhaps a desire that is instrumental to still another desire

and so on until we arrive at ultimate desires. Pain is one such ultimate desire because

it is something we desire to avoid for its own sake—it is intrinsically undesirable.

There of course may be desires that have both instrumental and ultimate dimensions.

For instance, we may want to avoid pain for its own sake, and we may also desire to

avoid pain in order to satisfy our desire for being productive. The desire to be

productive could itself trace back to a more ultimate desire to experience a kind of

pleasure, and thus the converse of pain, rendering a desire to be productive an

instrumental one in the end. What many philosophers want to deny, however, is that

the desire to avoid pain, or conversely the desire to seek pleasure, is the only ultimate

desire for human beings. In other words, they embrace genuine altruism. Those

whose works are situated in evolutionary frameworks, such as S&W, argue that a

pluralistic framework of ultimate desires has been selected for in humans, and an

altruistic desire is one such ultimate desire.

More generally, in the philosophical literature on altruism, and most notably

in Joseph Butler, C.D. Broad, Joel Feinberg, and Sober and Wilson, definitions of

altruistic motivations have been in terms of the kinds of desires they are and in

particular desires that are fundamentally distinct from egoistic desires; they do not

reduce to egoistic desires. Thomas Nagel’s view is distinct in that he does not define

altruism in terms of desires, but rather in terms of a sense of duty. However, in

Nagel’s account as in the other accounts, altruism is still defined by an ultimate as

opposed to an instrumental concern. Since the differences between a sense of duty

and a desire will not matter for my arguments, I will include Nagel’s account in the

28

standard view. In addition, I wish to briefly mention one particular work that does

not fit neatly into the standard view but is salient in the contemporary literature, and

this is The Heart of Altruism by Kristen Monroe.7 Monroe’s empirical work is

noteworthy for its distinct interview approach for understanding altruism. However,

there are several concerns with the way in which she characterizes altruism.8 First,

she identifies altruism with a perspective rather than merely a motivation; and this

seems too cognitively demanding to be an explanation for much altruistic behavior.

Second, this perspective is one in which the altruist understands the “common

humanity” of the potential victim; and this collapses meaningful distinctions between

altruism and abiding by normative principles. Finally, the perspective is an impartial

one; and this fails to allow that many of our altruistic motivations and behaviors are

about loved ones and those whom we know personally, whether our families,

cultures, or communities. The net result of Monroe’s conception of altruism, I

believe, is to raise the bar for what counts as altruism to be too high. Altruism

becomes a rare and demanding capacity that is limited to only a few unique

individuals, and does not get the chance to compete as a possible widespread

explanation for many of our actions.

Returning to the standard view, genuine altruism involves an ultimate desire

that is other-directed and egoism involves an ultimate desire that is directed towards

the self.9 But of course we need to say more about both of these desires. Taking

7 Monroe, Kristen Renwick. 1996. The Heart of Altruism: Perceptions of a Common Humanity. Princeton University Press. 8 See especially chapters 10 and 11. 9 S&W want to treat certain desires, such as Jack’s desire that a resource is split between himself and Mike, as a separate category from either egoism or altruism. They label desires that are both self and other directed relational desires.

29

altruism first, I could have a desire that is other-directed, but is also malicious. For

instance, I might have the desire that you lose all of your assets in a market crash.

Even though this is a desire directed towards someone other than myself, we would

not want to call this desire altruistic. Only desires to benefit others in some way can

rightly be called altruistic. Now taking egoism, we may not want to call, for instance,

a desire that oneself be tortured for its own sake an egoistic desire. It is more

appropriate to say that desires that aim at the benefit of the self in some way are

egoistic ones. Thus, altruism consists in an ultimate desire for the improved welfare

of another, and egoism in an ultimate desire for the improved welfare of the self.10

Egoism can be divided still into further categories by giving content to the

notion of welfare. For instance, welfare can consist either of any kind of external

good, such as material goods only or material goods combined with public approval,

distinction, or fame, or welfare can consist of internal rewards, and namely pleasure

and the avoidance of pain.11 Since arguments for the first category can be more easily

refuted by citing examples of an individual taking some action where there is no

material or recognition returns to be had, the strongest position for an egoist is

hedonism. Again, hedonism is the view that whatever action an agent performs, the

ultimate motivation of the action is to avoid pain or attain pleasure. Taking hedonism

as the strongest form of egoism, then, I will treat hedonism and egoism as equivalent

10 Egoism is silent on the question of whether individuals are motivated by short- or long- term interests or by both. Likewise, altruism is silent on the nature of the benefits to another—whether what constitutes something being beneficial can be construed as whatever I, as someone who has desires for your welfare, perceives as beneficial to you; or whether what is beneficial must be decided by the object of my desires. 11 Brennan and Petit argue for instance that the desire for esteem and status play a large role in both economic and political life. (2004) The Economy of Esteem: An Essay on Civil and Political Society. Oxford: Oxford University Press.)

30

except for certain occasions when it is relevant to distinguish them. For the bulk of

the discussion, we are left with the definition of egoistic motivation as an ultimate

desire for pleasurable feelings and altruistic motivation as an ultimate desire for the

improved welfare of others. As a general descriptive thesis, ‘egoism’ asserts that the

only ultimate desires we have are the desires for pleasurable feelings. The egoism

thesis does not maintain that we never have desires to help others—it is obvious that

we do, at least sometimes, have desires for the improved welfare of others. What

egoism specifically denies is that these desires for others’ welfare are ultimate. While

we can have instrumental desires for the welfare of others, the only ultimate desires

are to maximize pleasure and minimize pain. The thesis of ‘altruism’ however does

not maintain that the only ultimate desires we have are the desires for the welfare of

others—that would be implausible and easily refutable. Rather, altruism as a

descriptive thesis maintains that at least some of our ultimate desires are for the

welfare of others. Put this way, it is a relatively moderate thesis that does not purport

to account for all of our motivations. It does not maintain that we are all altruists all

the time or all of us are altruists some of the time or even that some of us are altruists

some of the time. It merely refutes the strong thesis of egoism which holds that there

are never any cases of altruistic ultimate desires. Indeed, Sober and Wilson’s

evolutionary arguments for the existence of genuine altruism only attempt to show

that some altruistic ultimate desires have evolved. I turn to this argument next.

31

1 .3 Evolutionary Arguments

In the first part of their book, S&W provide a strong case for the evolution of

biological altruism. Biological or behavioral altruism is defined as occurring when

an individual organism incurs a cost in the process of benefiting another. As

understood in the biological sense, altruism merely captures a type of behavior,

namely behavior that benefits another at an expense to oneself, and is silent on

whether or what kind of motivations are operating. However, S&W reason that some

internal mechanism must be responsible for getting the organism to behave in

evolutionarily advantageous ways. After arguing for the selection of biological

altruism by competition between groups, they argue that for organisms with minds,

such as humans, motivations, and namely desires, are what serve to achieve altruistic

behavior. What triggers or causes a desire is a belief, such as the belief that someone

has been harmed. As part of the standard view, S&W maintain that the combination

of belief and desire constitutes a complete motivation to act.

So what is left to determine is the nature of these desires since different

internal mechanisms can be functionally equivalent in terms of causing the same kind

of behavior. Indeed, the debate about genuine altruism take this problem to be

central—altruistic behavior does not yet tell us whether one is a genuine altruist or

not. Even if arguments and evidence can be mounted to support biological altruism

through group-selection, as S&W have done, it is still a far cry from demonstrating

the existence of motivational altruism. One can perform an altruistic behavior of

sacrificing her own fitness for the sake of (unrelated) others and, at the same time,

have done so because she experienced psychological pain from having observed the

32

others in harm’s way. The latter motivation would describe, according to the

standard view, an egoist. Thus, one can be both a biological altruist and a

motivational egoist. Furthermore, and most importantly, arguments advanced for the

existence of motivational altruism do not depend on the success of arguments made

on behalf of group-selection, which is still a contentious view in biology, or even on

the existence of biological altruism. Biological and motivational altruism are two

distinct phenomena and the domain of interest to philosophers and others interested in

the rational choice debate is motivational altruism.

Historically, egoism has enjoyed a privileged status as the default assumption

about motivations in the social sciences. One of the reasons for this seems to be that

the social sciences have looked to the sciences, and especially biology, to inform their

methodology, and egoism has been more widely accepted than views that also include

altruism as an explanation of behavior in biology. When it comes to motivations,

defenders of egoism argue that positing a single ultimate motivation is more

parsimonious. Altruistic motivations are viewed as simply being unnecessary clutter

in explaining our reasons for acting. This argument from parsimony is something like

the following: pain and pleasure do such a good job of regulating our behavior

because pain and pleasure are strongly correlated with harm and survival,

respectively, so there simply is no need to posit any other fundamental motive than

that of avoiding pain and seeking pleasure. What often happens when one is

attempting to explain an action in terms of an altruistic motivation is that defenders of

egoism are quick to respond with a plausible-sounding hedonistic explanation for the

33

action, and they claim they have parsimony on their side; this ‘hedonistic trap’ is

difficult for the defender of altruism to climb out of.

One of the reasons for the seeming ability of egoism to explain so much is its

flexibility—egoism does not tell us exactly what our motives are but rather tells us

what kinds of motives to look for, namely self-directed ones.12 So when one egoistic

explanation fails, another one can be readily offered in its place. But there is an even

more important, albeit related, source of flexibility for egoism, and for hedonism in

particular. Pain and pleasure are very vague and general psychological concepts.

Consider how broadly we think of pain. Painful feelings can be anything from simple

sensations to more complicated emotions, from an aversive sensation in one’s ear

from a loud sound to sophisticated emotions such as guilt. Pain can range from

‘passive’ sensations and feelings such as fatigue or shame to more ‘aggressive’ ones

such as hunger or disgust. It can be experienced as focused on another, as in

sympathy, or experienced as focused on the self, as in personal distress, and pain can

be coolly felt in the form of regret or can be passionately felt in the form of

resentment. Because of the vagueness and generality of the concepts of pain and

pleasure, hedonism, which holds that our single ultimate desire is to avoid pain and

seek pleasure, is also vague and general. The content of this ultimate desire is not a

single thing but includes, in the case of pain, anything we do not like, and in the case

of pleasure, anything we like.

Hedonism’s flexibility makes it far too easy for an egoist to insist that an

action was performed for hedonistic motivations whenever the actor cites any kind of

12 S&W, pp.288-289

34

feeling, emotion or sensation. It is not surprising therefore that philosophical

treatments of altruism have focused on carving strong boundaries between altruism

and hedonism in particular. Unfortunately, attempts to avoid falling into the

hedonistic trap have unnecessarily and to its detriment, I will argue, divorced feelings

from altruism.

Whatever we consider to be the greatest source of hedonism’s, and more

generally, of egoism’s flexibility, that it is flexible is part of the reason that egoists

can claim that it explains all of our behavior without having to posit any other

motives. Its flexibility is part and parcel of its so-called parsimony. But S&W

believe that the argument from parsimony is not a good one.13 They maintain that

merely counting the number of ultimate motives is insufficient for determining the

parsimony of a system, and furthermore, they remind us that parsimony is supposed

to be an all-else-equal tool of deciding between competing theories. If there are other

strong reasons, other than appealing to the standard of parsimony that is, for favoring

a theory that posits pluralistic motivations then these other reasons should be

considered and may end up tipping the scale in favor of altruism.

S&W argue that there are three main criteria other than parsimony by which

we should weigh competing theories about the evolution of certain strategies: (1)

reliability; strategies that stably contribute to fitness are more likely to have evolved,

(2) availability; only traits that are available can be subject to selection pressures, and

(3) efficiency; a trait which employs less of the organism’s energy resources has a

competitive advantage. S&W believe that the sum effect of these considerations is to

13 S&W, pp. 291-295

35

tip the scale in favor of altruism.14 As mentioned above, motivations can be

described as the internal mechanisms for causing a human being to behave in ways

that are selected for through natural selection. After arguing that selection has

favored some helping behavior towards others, S&W suggest that we need to explore

which motivations, and in particular desires, would have evolved in order to bring

about helping behavior. They describe this inquiry as determining the solution to the

design problem of altruistic behavior. I only wish to examine one of their standards—

reliability—because it is this standard that they argue determines the difference

between the success of altruism over egoism in explaining motivations for a particular

sort of behavior. Taking this standard of reliability, they argue that a pluralistic

framework that includes altruism as at least one type of ultimate desire will be more

reliable in bringing about altruistic behavior than a monistic framework consisting of

egoistic motivations alone.

To begin their argument about reliability, S&W make the very plausible

assumption that direct internal pathways will more reliably bring about a fitness-

enhancing behavior than indirect internal pathways, and they illustrate this point with

an example about bacteria. Marine bacteria must avoid oxygen in order to survive.

There are two kinds of mechanisms that could have evolved to allow a given

bacterium to detect the presence of oxygen. One would be a straightforward and

direct “oxygen-detector”. The other type would be a detector of some variable that is

correlated with oxygen, such as elevation in water. They call this indirect mechanism

a “toxygen-detector”. If toxygen and oxygen are perfectly correlated, then bacterium

14 pp. 297-327

36

will do just as well with either a toxygen- or an oxygen- detector. In fact, if the

oxygen-detector is somehow inaccurate or flawed at detecting oxygen, then the

toxygen-detector would be a more reliable guide for avoiding oxygen. But if the

oxygen-detector performs at least as accurately as the toxygen-detector plus, and this

is the important point, toxygen and oxygen are not perfectly correlated, then the

oxygen-detector is a more reliable pathway for avoiding oxygen. Given the same

level of performance of each mechanism, the direct detector (D) will be more reliable

than the indirect detector (I) in producing oxygen-avoiding behavior. S&W call this

conclusion the D/I Principle.

Turning to altruistic behavior, S&W assess which of two motivational

pathways—an egoistic ultimate desire or altruistic ultimate desire—is more reliable in

bringing about altruistic behavior. They argue that desires are a sort of instruction

that tell us what actions to take given what beliefs we have. They demonstrate the

contrast between altruistic ultimate desires and egoistic ultimate desires by

representing the different motivational pathways involved with bringing about a case

of altruistic behavior that is very widespread: parents helping their children. A parent

who only has egoistic desires will take any action only if she believes that the action

will maximize pleasure or minimize pain. A parent who only has altruistic desires

will take an action only if she believes the action will best serve her child’s welfare.

It is (plausibly) assumed that taking care of one’s children, and not merely producing

them, contributes to the reproductive success of the individual. If a child perishes

before he himself has the opportunity to reproduce, this limits the parent’s long-term

fitness. It is important to again stress the difference between motivational and

37

biological altruism. Because a parent (or rather her genetic material) receives

benefits from taking care of her child, it might seem as if such actions should not

really count as genuine altruism. But this would be to confuse the biological with the

motivational sense of altruism. Whatever the biological benefits of some behavior,

what S&W, other philosophers, and indeed commentators on altruism in most

disciplines aim to determine is the nature of the motivation behind those actions.

Again, genuine altruism in the motivational sense consists of an ultimate altruistic

desire and does not depend on whether there are benefits to the individual. The only

stipulation is that the individual is not being ultimately motivated by these benefits.

Analogous to the marine bacteria example, in the case of a parent helping her

child we need to determine which motivation, an egoistic one or an altruistic one, will

perform more accurately in producing the fitness-enhancing behavior. S&W suggest

that it is possible for parents to have flawed beliefs about what will best serve their

children’s welfare, even though they do not think it is the norm. But we could also

grant that the same can be said about the beliefs of the egoist. People may have

flawed beliefs concerning what will maximize their pleasure or minimize their pain.

In the absence of any evidence indicating which of the two kinds of beliefs are more

likely to be flawed, for now we should treat the two motivational pathways as having

the same level of performance.

What is left to determine is whether one of the motivational pathways is more

direct than the other. Reproducing their diagram, the two different motivational paths

look like the following, with the altruistic parent (A) represented in the first column,

and the egoistic parent (E) in the second:

38

Figure 1

Child needs help → Parent believes Child needs help

↓ ↓ ↓ ↓

A’s Desire for E Feels Bad that Child’s Welfare child needs help

↓ ↓ ↓ ↓

A Helps Child E’s Desire to Minimize Pain

↓ ↓ E Helps Child

When the A parent acquires the belief that her child needs help, she

formulates the desire for her child’s welfare. When the E parent acquires the belief

that her child needs help, she first feels bad and as a response to this bad feeling, she

formulates the desire for her child’s welfare. S&W argue that because E’s action to

help her child is mediated through her feelings, there is an extra step so to speak

involved when E helps her child than when A helps her child. They conclude that it

is clear that an altruistic desire to improve your child’s welfare is a more direct

mechanism for helping behavior than is an egoistic desire.

To relate this point back to S&W’s marine bacteria example, we can put this

point a slightly different way. The object of A’s desire is her child’s welfare, so she

can be said to have a “child-welfare detector”, whereas the object of E’s desire is to

minimize pain/maximize pleasure, so she can be said to have a “pain-detector”. Just

as toxygen must be correlated with oxygen in order for a toxygen-detector to be

successful in avoiding oxygen, E’s painful feelings must correlate with her child’s

welfare in order for her pain-detector to be successful in improving her child’s

welfare. Recall that S&W’s D/I Principle states that given the same level of accuracy

of two detection mechanisms, the direct one will be more reliable than the indirect

39

one in bringing about the fitness-enhancing behavior if, and again this is the

important qualification, what the indirect detection mechanism detects is less than

perfectly correlated with what the direct detection mechanism detects. In other

words, we now need to find out whether E’s pain/pleasure is less than perfectly

correlated with her child’s welfare.

S&W think there are two considerations that bear on the question of whether

pain and pleasure are perfectly correlated with welfare. The first point is that some

patients who have undergone a bodily injury have reported that they did not initially,

and perhaps even some time afterwards, feel pain. The second, and opposite, point is

that some individuals report the experience of pain in the absence of bodily harm.

Nevertheless “pain is an extremely useful, but imperfectly reliable, indicator of bodily

injury” when we are talking about someone’s own body. 15 It is a different story,

however, when it comes to someone detecting the pain in the body of her child. In

the sort of example being discussed, E’s pain must be correlated to her child’s welfare

and not her own. S&W assert that the only way this can happen is by way of beliefs.

When someone’s own body is injured, she feels pain directly without having to form

any beliefs about whether she was injured, and so pain-detection is a direct

mechanism for producing behavior that improves her own welfare. However, when

someone’s child is injured, she feels pain only after she has obtained the belief that

her child is injured. So E’s pain will be even less perfectly correlated with her child’s

welfare, because it is mediated by a belief, than it is with her own welfare. The

conclusion that S&W draw is that since A’s motivational pathway is a more direct

15 S&W, p.316

40

mechanism for helping her child—it is not mediated via feelings—and since E’s pain

is not perfectly correlated with her child’s welfare, A’s motivational pathway is more

reliable than E’s motivational pathway. S&W take this to mean that as an

evolutionary strategy to solve the design problem of caring for one’s children,

altruistic ultimate desires are more reliable than egoistic ultimate desires. Thus, their

general conclusion is that there are strong reasons to believe that genuine altruism has

evolved and exists.

1 .4 Problems with Evolutionary Arguments for Genuine Altruism

Despite the seeming plausibility of S&W’s arguments, and the optimistic

conclusion they draw, there are several problems with their arguments. The first point

I want to make concerns the role of beliefs in the egoism pathway. S&W argue that

desires operate in the example above only through a belief that is relevant to the

object of desire. In order for E to feel bad when her child needs help, she must first

form the belief that her child needs help. Because of the role of this belief, a case of a

parent responding to the harm to her child is very different from a case of someone

responding to her own bodily harm. When I injure my leg, I do not need to first form

a belief that I have injured my leg in order to feel pain. I feel pain as a direct result of

what is happening to my body. E’s painful feelings about her child, S&W argue, are

not nearly as direct.

Contrary to what S&W maintain, however, it seems that there are at least

some circumstances when E’s painful feelings about what is happening to her child

could occur in the absence of forming any belief. Consider when someone you know,

41

or even someone you do not know, is performing in some way—perhaps dancing on

stage or delivering a presentation. If this person looks like he is about to trip and fall

in the middle of his performance but then stabilizes, you might have a tingling

sensation in your gut, and a quickening of your heartbeat some time after his near-

fall, despite the fact that you do not have the belief that he fell or will fall. There are

better examples. Think about a movie where a character is being tortured by a

malicious killer. Even if you consciously think to yourself that what appears to be

happening is not in fact happening, you might actually experience pain in the parts of

your body that correspond to where the character is being tortured. It would seem odd

if I were to insist that you must be having the belief that this person on your screen is

actually being tortured or has been tortured.

One might think that the dissimilarity between pain experienced in our bodies

due to our own bodily harm and the possible pain experienced in our bodies in

response to the bodily harm in others is still too strong. After all, our immediacy with

our own bodies’ pain is what grants pain its superior epistemic status—we are wired

to experience the pain of our own bodies for the purpose of informing us of bodily

harm. But if we broaden our notion of what constitutes pain, the difference seems to

go away. As S&W themselves acknowledge, egoism does not hold that the category

of pain just applies to our physical sensations. Negative emotions, feelings, and

perhaps even moods, if they are sufficiently aversive, count as pains. As I mentioned

above, this broadening of the notion of pain is what makes egoism, and in particular

hedonism, such a flexible view. But once we include emotions and feelings, it seems

more obvious that there are occasions in which we experience pain in response to the

42

suffering of others without having to first form beliefs. How many times have you

cried by watching sad movies or by reading sad novels which are fiction? You might

insist that you are forming some belief, perhaps that the plight of the character

resonates because you are aware that others are experiencing that plight in the real

world. Or perhaps, we are really being duped into believing that these actors on the

screen are suffering, if only for a moment.

We should be skeptical of these responses, but even if one holds their ground,

there is evidence of ‘emotion contagion’ even when the circumstances do not involve

fictional characters. For instance, social-psychologists have studied people’s facial

expressions as participants observe the expressions of others and found that there are

different response patterns, as measured by electromyographic (EMG) analysis, when

participants observed happy versus angry faces.16 And the connection between facial

expressions and the experience of emotions appears to be direct and specific. Paul

Ekman and his colleagues had participants either recall and try to relive experiences

of a certain emotional kind, such as anger or sadness, or, arrange their facial muscles

in ways that are characteristically associated with these emotions. The researchers

found that both groups of participants experienced the autonomic nervous system

(ANS) responses that are typically associated with the particular emotion.17 Emotion

contagion frequently appears to occur automatically, without the mediation of belief

or other conscious processes. Mark Davis argues that the synchrony of a multitude of

facial, vocal and postural responses is too fast and too extensive to always be

16 Dimberg, Ulf. (1982). “Facial reactions to facial expressions.” Psychophysiology, 19, 643-647. 17 Ekman, P., Levenson, R.W., & Friesen, W.V. (1983). Autonomic nervous system activity distinguishes emotions, Science, 221, 1208-1210.

43

mediated through conscious processes.18 He argues that it probably involves

multiple-levels and areas of the brain simultaneously. In one study, participants

mimicked and synchronized the postures and movements of others within 21

milliseconds.19 That’s a pretty speedy rate when you consider that in practice

sessions it took Muhammed Ali at least 190 milliseconds to detect a light and 40

more milliseconds to respond by throwing a punch—230 milliseconds for something

that we might reasonably put in the category of cases of a person reacting to threats to

his own bodily injury, and not to that of others.20

Maybe the best example of emotion contagion happens within the first few

days of our lives. Consider how frequent it is for newborn babies to cry when other

babies are crying. Whether crying-contagion is a form of empathy or sympathy or

other complex emotions, or even any emotion at all, is a matter of some controversy.

But it seems that babies are experiencing some distress, however self-focused, when

they cry in response to other babies crying. Distress is an aversive feeling, and the

hedonist has no problem accepting that babies feel distressed simply by perceiving

the distress of others without first forming a belief.

If babies experience aversive feelings simply by registering the distress of

other, unrelated babies, it seems just as plausible that a baby’s mother, or father, is

capable of experiencing aversive feelings when her, or his, own baby cries simply by

18 Davis, Mark. H. (1985). Perceptual and affective reverberation components. In A.B. Goldstein and G.Y Michaels (Eds.), Empathy: Development, training, and consequences (pp. 62-108). Hillsdale, NJ: Erlbaum. 19 Condon, W.S., & Ogston, W.C. (1966). “Sound film analysis of normal and pathological behavior patterns.” Journal Nervous Mental Disorders, 143, 338-347. 20 The example comes from Hatfield, Elaine and Rapson, Richard L. (2004) Emotional Contagion. In Larissa Z. Tiedens and Colin Wayne Leach (Eds.), The Social Life of Emotions (pp.129-43). Cambridge, UK: Cambridge University Press.

44

registering her babies’ distress. Let us take a look at how the egoistic (E) parent’s

motivational pathway might be different if her feelings did not have to be mediated

by beliefs.

Figure 2

Child needs help ↓ ↓ A’s beliefs E’s feelings about Child about Child

↓ ↓ ↓ ↓

A’s Desire for E’s Desire to Child’s Welfare Minimize Pain

↓ ↓ ↓ ↓

A Helps Child E Helps Child

I have presented E’s feelings as occurring directly in response to observing

that her child needs help, and E’s desire is mediated through her feelings.

Furthermore, precisely because the view of altruism under discussion is one that does

not require feelings or sensations, A’s desire is not presented as being mediated by

feelings or sensations, but by a belief. If E’s motivational pathway operates in this

way, then notice that it takes E just as many steps as it takes A to help their respective

children. In other words, A is not a more direct solution to the problem of helping

behavior in parents, and without further argument, A is not more reliable than E.

None of what I have just discussed implies that all of E’s pain for her child occurs in

the absence of beliefs. If E is in a different country than her child and hears from a

third party that her child is injured, E’s aversive feelings might very well only occur

by first forming a belief about her child. Nevertheless it seems that some, possibly

45

many, of E’s feelings about her child’s welfare could occur without the prior

formulation of a belief.

In the first part of their book, S&W argue that altruistic behavior or biological

altruism can be selected for because of its advantageousness for one’s kin or group.

Likewise, having the capacity to immediately and directly feel pain in response to

perceiving that others are in pain could plausibly be selected for because of the same

reason—because it is advantageous for one’s kin or group. It is efficient for an

organism to feel pain when her hand is burning and withdraw her hand without her

first having to formulate a belief that her hand is burning. Just as it would be

inefficient for her withdrawing of her hand to be keyed to her beliefs, it would also be

inefficient, from the standpoint of caring for one’s children, for a parent to remove

her child’s hand from a fire only after forming a belief. My aim here is not to argue

one way or the other that pain in response to another’s pain has been selected for or

that there is an evolutionary advantage to such a response. My aim is only to point

out an alternative understanding of pain response that would not be ruled out, and

could be supported, by the type of evolutionary argument that S&W advance. More

generally, it is to convey a problem with relying on evolutionary arguments to support

the existence of altruism. Aside from the possibility of an evolutionary argument for

the existence of some trait being a “just-so” story, there is a further problem more

specific to the phenomenon of genuine altruism. I will turn to this more serious

problem with S&W’s argument next.

Recall S&W’s D/I Principle as applied to altruistic and egoistic motives. It

states that given the same level of accuracy of the two motivational pathways, the

46

direct one will be more reliable than the indirect one in causing parents to take care of

their children if there is a less than perfect correlation between what the indirect

motivation is keyed into and a child’s welfare. A desire for the improvement of

another’s welfare is a direct internal mechanism for causing helping behavior and a

desire to avoid pain or seek pleasure is an indirect mechanism. Assuming S&W are

correct in arguing that feeling pain in response to another’s pain only comes after

belief-formation, then the reason for the indirectness in the egoism pathway is the

additional step it involves. Altruistic desires are therefore more reliable in promoting

other-directed behavior than egoistic desires.

The problem with S&W’s argument, however, is that the satisfaction of the

D/I Principle only requires an altruistic desire. It is silent on the nature of this desire.

In particular, it does not tell us whether the desire has to be an instrumental or an

ultimate altruistic desire. In fact, if A’s desire for the improvement of her child’s

welfare were an instrumental desire, this would satisfy the D/I criteria. It could be a

desire that serves another more ultimate desire, however near or remote, and because

the role this ultimate desire plays in A’s motivation is to bring about the existence of

other desires, it operates at a step removed. It is not directly tapped into, so to speak,

when A’s child needs help. Instead A’s desire to help her child is what gets activated.

This is not to suggest that A’s ultimate desire would be an unconscious one—

although, as I will discuss below, S&W think even this is perfectly possible and

should not be ruled out. It only suggests that A’s ultimate desire would not be

experienced at that time, whereas E would experience her ultimate desire to avoid

pain—this is what would account for the indirectness of E’s motivational pathway.

47

But, and this is the important point, even though A’s desire to help her child is more

direct than E’s desire to avoid pain, A’s desire would still only be an instrumental

one. Presumably, A’s desire would be relatively more ultimate than E’s desire, since,

again, E would also experience the desire to avoid pain. However, on the view that

genuine altruism involves an ultimate altruistic desire, the fact that A’s desire to

protect her child would be relatively more ultimate than E’s does not yet tell us

whether A is a genuine altruist or not.

A might have desires for many things that we would want to call altruistic,

each with a different object, such as to promote peace, to read literature to

disadvantaged youth, to look out for her friends, and to promote the wellbeing of her

child. But as long as these desires only exist as means to bring about the satisfaction

of other, more fundamental, desires, they are all instrumental desires. In fact, she

could, ultimately, have only the single desire to minimize pain and maximize

pleasure.

This observation poses a serious problem for S&W’s effort to establish the

evolutionary reliability of genuine altruism since their definition of genuine altruism

entails an ultimate, and not an instrumental, desire. If an ultimate desire to minimize

pain/maximize pleasure is what gives rise to A’s desire to protect her children, then A

is not an altruist and S&W have not shown us why it is plausible to think that genuine

altruism has evolved. More generally, the difficulties facing any evolutionary

argument for the existence of genuine altruism should now be clear. Despite being

convincing arguments for the emergence of altruistic motivation, evolutionary

arguments such as S&W’s could not hope to show us why this altruistic motivation

48

would come in the form of ultimate and not instrumental desires, since, as I have

demonstrated with the alternative interpretation of A’s desires, the two are

functionally equivalent. Instrumental desires serve as reliably as ultimate ones in

producing altruistic behaviors when we, plausibly, maintain that ultimate desires can

be background ones that are not always experienced at a given time.

1 .5 Conclusion

So where does this leave us? S&W offer the most convincing argument to

date in support of the notion that genuine altruism has evolved, and more generally,

that it exists. Since their arguments fall short, does this mean we are doomed to

believe that there are no good arguments demonstrating the existence of genuine

altruism and so we must believe that we are all, ultimately, egoists? I do not think so.

Arguments such as S&W’s fail to be convincing because they employ the standard

view of genuine altruism, which forces us to parse out ultimate from instrumental

desires. I will argue, however, that when we reflect a little more on the nature of

altruism, the view that the genuineness of altruism depends on the level of desire—or

whether a desire is ultimate—is not a compelling view of altruism. In other words, it

is not the case that in order to argue for the existence of genuine altruism we will

have to accept a less accurate notion of what we mean by altruism.

It seems the greatest strength of S&W’s argument for the evolution of

motivational altruism has to do with the flexibility of a system in which altruism

would be embedded. A pluralistic motivation system is one that would include both

egoistic and altruistic motivations. S&W reason that pluralism with respect to

49

motivations is more plausible, from an evolutionary perspective, than monism. Kim

Sterelny has made a similar argument about the benefits of plasticity that come from

motivational pluralism.21 A pluralistic parent, one with both altruistic and egoistic

desires, will more reliably help her children than a purely egoistic parent. Whereas

the egoistic parent only helps her child when it minimizes pain, the pluralistic parent

will help her child when it minimizes pain or when it promotes her child’s welfare. If

one motivational pathway fails, there is another one to do the job. In some cases,

helping her child will both minimize pain and promote her child’s welfare. So she

will have two reasons to help her child. As long as the two pathways operate

independently so that they do not interfere with each other, the two pathways are

more reliable than either pathway working alone.22

The evolutionary arguments for motivational pluralism in cognitively

sophisticated beings such as humans are strong, but it is not necessary that the kind of

pluralism that has evolved pertains to different ultimate desires. The same arguments

support the evolution of pluralism with respect to different emotions or feelings,

which, I will argue next, are a distinct kind of motivation from desires. Since I will

argue that emotions rather than ultimate desires play a prominent role in our

conception of genuine altruism, it is important to spend some time, in the following

Appendix, saying what emotions are and how they differ from desires. In addition,

21 Sterelny, Kim. 2003. Thought in a Hostile World: The Evolution of Human Cognition. Blackwell Publishing 22 I can think of cases where the two pathways might conflict. I might feel bad punishing my child for something she has done but at the same time believe that it is promotes her welfare in the long-run. How would I choose? Maybe a purely egoistic parent would face similar conflicts. I might feel bad punishing my child but at the same time feel pleasure from the pride of being a firm parent. While comparability might be somewhat less complicated for a pure egoist, the argument from pluralism is a good one.

50

my analysis of the implications of my view of genuine altruism in the Appendix to

Chapter 2 builds off of this discussion on the emotions. However, if the reader

chooses to skip this section for now and go directly to the discussion of the

philosophical arguments for the existence of genuine altruism, she may proceed to the

beginning of Chapter 2.

51

1 .6 Appendix 1 : Are Feelings and Desires Separate?

In Gut Reactions, Jesse Prinz provides a discussion of the varying views on

emotions, focusing on all dimensions important to each prominent view. In order to

sort through them, I begin by organizing the kinds of theories around the two

dimensions that matter most to my discussion here: 1) motivation; because the debate

between genuine altruism and egoism is a debate about what motivates our behavior;

and, 2) feeling; because it is this dimension that I will argue is missing from the

standard view of altruism. Below is a representation of the typology of emotion

views as they pertain to these two dimensions.

Table 1

No Motivation Motivation

No Feeling 1: Cognitive Views 2: Behaviorist View

Feeling 3: Somatic Views 4: Hume

According to Prinz, many theories of emotions maintain that emotions have

motivational force. For instance, a variety of views hold emotions to be action-

tendencies. At one end of this spectrum are certain behaviorist views, represented in

cell 2, where emotions are only dispositions to act in certain ways. In other words,

emotions are described in the same way that many philosophers describe desires.

These same behaviorist views also maintain that there is no feeling element to

emotions. Some views divorce emotions from both feelings and motivations. For

instance, Cell 1 contains cognitive views, which hold that emotions are judgments

like beliefs, such as a belief that one has been wronged, or are judgments about

judgments, such as an evaluation that the belief that one has been wronged is

52

justified. Both the Cognitive and the Behaviorist views lack what seems to be

intuitively central to emotions, that emotions involve feelings. In one study, every

group of people surveyed except one rated the feeling component of emotions to be

the most important.23 Whether or not feelings accurately capture the essence of every

emotion, it is intuitive that feelings are central to many. In any case, feelings of a

certain type will end up being important for the discussion of altruism, and therefore

so will views of emotions that give a central role to feelings. There are two kinds of

views that do this.

At the intersection of both the motivation and the feeling dimension is

something like Hume’s view, where emotions are feelings that contain desires, and

desires according to Prinz are themselves a subset of feelings that impel us to act.

Prinz writes that “fear may contain the felt desire to flee." A desire to flee is, for

Hume, an action tendency.”24 A great deal has been written about the nature of desire

in Hume. That all desires for Hume have a phenomenological feel is a matter of

dispute. Michael Smith interprets Hume as maintaining that only some desires are

experienced as feelings.25 He denies that Hume’s desires are only or even always

feelings. More broadly, even aside from his reading of Hume, Smith makes a

compelling case for the implausibility of viewing all desires as having a

phenomenological feel. He asks us to consider a long-term desire, for instance, the

long-term desire to be a doctor. In moments of reaffirmation of this desire, following

23 Jaak Panksepp, 2000. Subjects were asked to rank possible emotion components such as action-tendencies, feelings, thoughts and bodily changes. The group that failed to rate feelings as most important were college students majoring in philosophy. This group rated thoughts to be most important. 24 Prinz, p. 11 25 For a good discussion of the different views of desires, see: Smith, Michael. 1994. The Moral Problem. Oxford: Blackwell Publishing. pp.111-119

53

periods of doubt, this desire to practice medicine might express itself with a kind of

force and perhaps even a feeling. But even in the absence of this feeling, during times

in which the desire is not being challenged or reassessed, we would not want to say

that this desire simply goes away. Furthermore, when someone is sleeping we would

not say that their desires simply cease. However, as Smith points out, feelings may

turn out to be central to certain desires, such as a desire to eat. In the dispositional

account of desires presented above, these desires would include a disposition to feel a

certain way in addition to a disposition to act in particular ways.

While desires do not always involve feelings for Hume, according to Smith,

Prinz argues that feelings are central to Hume’s account of the emotions. But they

may contain more than this. Emotions for Hume also seem to contain cognitions

insofar as thoughts are used to distinguish one emotion from another. If this is the

case, then Hume’s view is also partly a cognitive theory. But, as Prinz has

convincingly suggested, we should not accept cognitive theories of emotions. A

moment’s reflection seems to dispel the notion that cognitions are defining aspects of

emotions. He offers the following example to support his point. “A friend has

convinced you to go on a hike. You are not much for the outdoors because you suffer

from a terrible snake phobia. Reluctantly, you go along. As you set out, all concerns

quickly drift away. You are having a good time. But then, out of the corner of your

eye, you glimpse a long coiled object nestled in the shadows. You freeze, your heart

pounds, your breathing becomes strained, your eyes widen.”26 Prinz’s point is that

despite not thinking about snakes, not having any cognitions related to snakes

26 Prinz, p. 21

54

whatsoever, you can imagine yourself becoming fearful the moment you spot

something that looks like a snake; even, perhaps, continue to be afraid for a few

moments after you acquire the belief that the object is not a snake. Prinz does not

argue that cognitions never cause emotions—he certainly believes they can—he just

maintains that cognition is not a necessary cause of emotion, and that cognitions are

not a part of emotions. This brings me to the final type of view, a view that Prinz has

called the “somatic feeling theory”.

Beginning with William James and Karl Lange, somatic theorists have

maintained that emotions respond to, register, or track changes in the body. Each

theorist has his own take on the nature of this response. For instance, for Antonio

Damasio, feelings are representations of bodily changes. If an organism is in the

presence of a predator, a harmful situation, its body will respond with an increased

heart rate, blood-flow and skin conductance, and the organism will begin to flee.

These bodily changes help organisms navigate the world by responding successfully

to stimuli. But then what work do feelings do? Feelings enter into the picture by

allowing the organism to be aware of these bodily changes and by giving the

organism an incentive to act. Feelings also help an organism use combined responses

to past situations in order to form novel responses for situations that have not

happened yet. 27

Although Damasio argues that feelings are important for their role in

representing bodily changes, he maintains that emotions need not involve conscious 27 For Damasio, emotional feelings contribute to another step in the development of sophisticated cognition: the feeling of knowing that we have feelings. What is the feeling of knowing? This is core consciousness. Damasio argues that consciousness evolved because it allowed organisms to feel feelings, or, in other words, to be aware of their emotions.

55

experiences. The brain can perceive bodily changes without causing feelings.

Emotions simply are the bodily changes that respond to stimuli in the environment,

and the feelings they give rise to are separate. To echo an earlier comment, what we

choose to include under the specific label of emotions is not terribly important for my

purposes. What I am interested in is this theory and others where feelings are

importantly linked to emotions, whether causally or constitutively. It happens to be

the case, however, that for Prinz’s theory of emotions, feelings also fall under the

label of emotions and are exhaustive of emotions. For this reason and for more

substantive reasons, his is the theory I favor. However I will offer certain

amendments to his theory.

Prinz argues that emotions are feelings that appraise certain situations, and

they do this by tracking bodily states. It’s more complicated than Damasio’s view

where emotions simply represent changes in the body, and so it deserves to be fleshed

out. Prinz argues that for emotions to represent something, they must be reliably

caused by that thing. Above, I briefly demonstrated how Prinz successfully argues

for the somatic view, so it would seem obvious that what reliably causes emotions in

Prinz’s view is the body, and that what emotions represent are bodily states.

However, Prinz also argues that for something to represent something else, it must

have the function of representing that something else, and it seems implausible that

emotions have the function of representing bodily changes.

If for instance I site a snake, my heart races, I feel afraid, and I run. Prinz

argues that it would be a mistake to think that my feeling of being afraid represents

my heart beat. My emotions represent something, but not, contrary to Damasio,

56

bodily changes. What would the survival advantage of that be? Prinz maintains that

emotions have the function of representing real situations. To see this, we can isolate

each, the body and the situation, to discover what it makes more sense for evolution

to have settled upon. Taking the body first, if feelings were to have the function of

representing bodily changes, then it is plausible that I would feel afraid, and flee,

simply if my heart beats faster, without there being any real situation that ought to

cause me to flee. That does not seem to make much sense. However, what if I feel

afraid, and flee, in the absence of any heart-rate increase, but because I confront a

situation that ought to cause me to flee? Now that would appear to be advantageous.

So what sort of situation ought to cause me to flee according to Prinz? One where

there is, or might be, a real danger present, like a snake. This kind of situation is,

more generally, “a relation between an organism and its environment that bears on

wellbeing”.28 Emotions are feelings that represent relations between an organism and

its environments that bear on its wellbeing. For instance, as already indicated, fear

represents danger to the organism and sadness represents a loss to the organism, and

the organism registers danger and loss through bodily changes.

To return to the point of my discussion on the emotions, I was exploring the

role of motivation and feelings in the emotions. Prinz’ view renders feelings

essential. But what about motivation? In addition to representing organism-

environment relations that bear on wellbeing, there is another component to emotions

in the Prinz view—valence. Valence is the intuitive idea that every emotion is either

a negative or positive emotion, either agreeable or disagreeable. Sadness, anger, and

28 P. 51

57

fear are negative emotions, while happiness and joy are positive, and certain emotions

such as surprise might be mixed. But according to Prinz, emotions do not have

negative or positive valence in virtue of the way they feel to us, for valence could be

an unconscious aspect to emotion.29 While I think this is debatable, what is more

relevant to my discussion on altruism is what Prinz takes to be the defining feature of

valence. For Prinz, the valence of an emotion is like an inner command to either want

more or less of a situation, all things considered. I have added the “all-things-

considered” qualification because, as Prinz notes, there may be other considerations

that outweigh the command. Echoing Damasio’s position described above, the

valence aspect of emotions is more like an incentive: emotions give us an incentive to

act. To sum up Prinz’s full view of emotions then, and adding the valence component,

the feeling of fear would not only represent a dangerous situation, again, through

registering changes in the body, but it would also represent the danger as something

that is bad or negative, and serve as an inner reinforcer to withdraw from the

dangerous object; to eliminate the situation.

Because it urges us to want more or less of an object or situation, valence

sounds an awful lot like a motivation. Prinz is sympathetic to this. However he thinks

there is a useful distinction to draw. Several emotions, he suggests, are not always

action-tendencies. For instance, he argues that anger does not always motivate us to

seek revenge and fear does not always impel us to flee.30 He notes that all emotions,

due to their valence marker, provide a reason for acting, but do not always in

29 Prinz doesn’t deny that emotions feel unpleasant or pleasant. Anger is a disagreeable emotion. It feels unpleasant. He just denies that there is one general feeling of unpleasantness or pleasantness that characterizes all negative or positive emotions respectively. 30 p. 193

58

themselves impel us to act. This is the essential difference between emotion and

motivation. When we are angry, the valence marker issues its command, “less of

this!”, but it is a command directed toward an inner condition (an angry mental state)

but it is not a command for action. Emotions facilitate, or prepare us for, action,

through bodily changes. But emotions do not consist of decisions, plans, or choices

to act, which have external situations and states as their object—that is the realm of

motivation. Prinz does believe the link between motivation and emotions is tight: he

denies that every action we take is a motivation—consider autonomic responses such

as my moving my fingers while I type—and wants to save the term motivation only

for responses that follow affective states, where “hedonic considerations” apply.

Since emotions are affective states, a motivation will frequently result from an

emotional state. But other affective states, such as hunger or fatigue, can trigger a

motivation and some emotions, again such as anger, will not always incite a

motivation. The two concepts can come apart.

I disagree with Prinz’s assessment here. There seem to be different senses of

motivation, and thus room to view the emotions as motivations of a certain type.

Furthermore, Prinz’s denial that emotions are motivations seems to stem from his

conflation of at least two of these senses of motivation—motivation as force, and

motivation as decision-making. More specifically, he employs different notions of

“action-tendency”, which is what he takes to be central to motivation. On the one

hand, Prinz seems to hold the view that action-tendency is like a force. It is difficult

59

to explain this notion well, but it is a familiar intuition.31 It is common to speak of

motivation in terms of having an “urge” or “drive”, or being “impelled” or “pushed”

to act. This seems to be the idea of action-tendency behind Prinz’s statement that

hunger, which he defines as a motivation, “commands us to eat” and is an “impulse to

eat” and his view that motivations are always triggered by affective states. 32 On the

other hand, Prinz also identifies action-tendency with terms such as decision and

choice, suggesting something that is more active than a mere tendency. For it seems

not only that many decisions and choices lack any kind of force of their own, but also

that they are realized subsequent to a tendency. For instance, when I decide to pick

up my tea cup, it would be more accurate to describe my wanting tea as a disposition

and my deciding to pick up my tea cup as what follows from my disposition. 33 It is

the course of action I choose in order to bring about the satisfaction of my want.34 A

decision to pick up my tea cup and take a sip or a choice to change the television

channel seem to have a stronger kinship with behavior and action than with a

disposition or tendency for action—on the assumption that tendency or disposition for

action is causally prior to action.

When Prinz denies that emotions are motivations he employs this sense of

motivation as decision-making. He argues that in the case of the emotion anger, for

31 Perhaps the difference between the two views of motivation can be brought out by noting what is implied by the passive versus the active voice. A motivation seems to be more like Prinz’s use of the term motive—a reason for action. But to say that something is motivating seems to imply that it is compelling you to act. 32 Ibid, p. 194-5 33 As Prinz notes, most of these terms—decision, want, desire, goal, hope, choice, etc.—are multiple-meaning terms. Again, it is not the exact term we employ, but the spirit of the term that seems to matter. 34 Notice that this doesn’t seem to be the same thing as my instrumental desire to pick up my tea cup, which I would maintain is a motivation.

60

instance, our bodies are prepared for action, but “at this point in processing, no action

has been selected, no strategy has been determined, no plan has been conceived. The

somatic state and valence marker must be fed into a mental system that selects

responses. Among the available responses is violent revenge against the source of

our anger…Once that choice has been made, we can say there is action tendency at

work.”35 If you view motivation as decision-making, then it is not surprising that

emotions under Prinz’s view would not count as motivations. For in the very

plausible somatic account of emotions that Prinz develops, emotions are bodily

feelings that merely represent, and in a sense respond to, organism-environment

relations, and this hardly seems to imply decision-making. But the decision-making

view of motivation is very different from a view of motivation as an urge, force or

push or pull. The two seem to be at odds with each other, for one involves a picture of

an agent as passive and as being acted on, and the other pictures the agent as active.

Rather than reconciling the two notions, it seems plausible to maintain that

there are different kinds of motivation. This point can be stressed by noting that

neither of the two accounts seems complete on its own. I have already discussed how

the decision-making account maintains a condition that seems too active for

everything we would want to call motivation. The force-view does not seem

completely right either. If action-tendency is a force then it seems that many of our

motivation terms, and many of the ways in which we explain or make sense of

someone’s actions, would not qualify as motivation. For instance, I currently have a

want to continue typing, but it would be misleading to think of this want in terms of a

35 Ibid., p. 194

61

“force” that is triggered by an affective state. Perhaps this want is instrumental to a

more ultimate want, such as wanting to finish my dissertation, for which the notion of

force or being impelled would be appropriate as a description. Nevertheless, I think it

is plausible to view my, albeit instrumental, want to type in this phenomenologically

weaker way. But how would you describe my wanting to continue typing if not as a

motivation? It seems we should allow for other kinds of motivations than only

affectively situated ones.

The insistence that there is only one sense of motivation fails to capture our

ordinary usage of a variety of motivation terms. Perhaps certain motivation is more

like decision-making, and another type is more akin to the non-affectively charged

desires and wants that are necessary to make sense of or to explain behavior, and still

another kind feels like a force, one over which the agent seems to have little control.

It seems emotions would qualify as motivations of this last type. Let us revisit

Prinz’s comments regarding forces or urges. He argued that emotions do not always

urge us to act, and for support he cites the occasional anger that fails to impel us to

seek revenge and fear that sometimes does not urge us to flee. In fact, I would

maintain that this is frequently the case with anger and fear. Most of the time I am

angry, I exhale loudly and curse, and these actions are a long way from seeking

revenge. But, contrary to decisions or choices, urges and forces do not seem to have

specific actions as objects. They are commands to do something in response, be it

kick a shoe or exclaim rage or to huff and puff. However, they need not always

compel us to take a particular action, such as seeking revenge.

62

Furthermore, unlike urges such as hunger or fatigue, any given emotion

represents a variety of different organism-environment relations, even if only subtle

variations on the same kinds of theme, and so the behavior that emotions command

will vary in response. To use anger as the example again, Prinz argues along with

several other emotion researchers that the kind of thing anger represents is an offense

to me or to those close to me. This seems correct, but offenses can come in different

forms and in different degrees, and these differences appear to have an effect on the

behaviors that anger commands. Hunger represents something quite concrete and

unvarying, namely, undernourishment, and no matter the degree of hunger, the action

command is the same: to eat. Anger, however, can represent varying kinds of

organism-environment relations, with more or less offensiveness and distinct kinds of

offenses—I can imagine being so angry at someone that I am compelled to avoid him.

It is not surprising therefore that anger sometimes fails to urge us to seek revenge.

Nonetheless, it compels us to do something.

To sum up my discussion about the relationship between emotions and

desires, I conclude that emotions are feelings and motivations, but they count as

motivations of only a particular kind—as a force or urge for action. While there is this

motivational force to emotions, this does not entail the more robust view that

emotions contain desires. Desires and emotions, or at the very least, feelings, seem to

be distinct. As we will see in the next chapter, this means that the standard view of

genuine altruism makes no essential reference to feelings of any kind. I will argue

that this is a significant shortcoming of the view. Furthermore, once we add feelings

to the picture, it becomes apparent that genuine altruism is more widespread than has

63

been traditionally assumed. Finally, in the second Appendix below, I will demonstrate

how including feelings as motivations as part of the conception of genuine altruism

may open up several promising avenues for both theoretical and empirical research.

64

Part I Rational Choice and Altruism

2 Ultimate or Instrumental Altruism? Who Cares

2.1 Introduction

I argued in the previous chapter that evolutionary strategies for demonstrating

that genuine altruism exists are problematic, and ultimately, fail to be convincing. In

this chapter, I will now turn to general philosophical arguments on the subject and

show why they too fail. I will argue that the source of the failure to establish the

existence of genuine altruism is the way in which genuine altruism is defined. As

stated in Chapter 1, the standard philosophical conception of altruistic motivation is

that it is a desire for the improved welfare of another, and genuine altruistic

motivation is an ultimate or non–reducible desire. Again, what triggers or causes an

altruistic desire is a belief, such as the belief that someone has been harmed. In a

similar manner, we could describe many instances of desires that are triggered

directly by beliefs alone. For instance, my desire to wear a raincoat is keyed directly

to the belief that it is raining outside. While it might be perfectly accurate to describe

the desire to wear a raincoat as not involving, and never involving, feelings, is it

accurate to describe altruism in this way, as occurring in the absence of feelings,

without any role for feelings? I do not think it is.

Historically, altruism has been associated with certain passions. Aristotle’s

description of filia, usually translated as “friendship” or “love,” is the passion that

describes the condition of sharing another’s pleasures and pains and he also believes

that good men feel pity for those who experience undeserved suffering. Both Adam

65

Smith and David Hume of course recognized our tendencies for fellow-feeling and

for perceiving the suffering of others through our ability to sympathize.1 Joseph

Butler, who wrote extensively on the topic, identified altruism, or benevolence, with

certain affections and love.2 Yet, despite some associations between altruism and

certain feelings, the standard view of altruism makes no essential reference to

feelings, and thus underplays their significance to altruism. In this chapter, I will

argue that the standard view does not present a sufficient understanding of genuine

altruistic motivation. In particular, whether a desire is ultimate may be an irrelevant

question for two reasons: first, a desire might not be ultimate but might be genuine

nonetheless; and, second, feelings are just as, if not more, important to the conception

of genuine altruism.

2.2 Prior Suitabil ity

To begin my analysis of the standard view, I want to introduce a major

philosophical argument against the thesis of egoism. Some authors, including S&W,

are unimpressed with this argument against egoism. But as I will suggest later in the

chapter, this argument may be revitalized once we consider the role of feelings. A

line of philosophers, starting with Joseph Butler and continuing with C.D. Broad, Joel

Feinberg, and Thomas Nagel, argue that many apparently hedonistic desires are not

egoistic upon closer analysis. They argue that the only way we can gain pleasure from

1 It is often argued that their notion of sympathy is much more akin to the way we use the term empathy today, as connoting a capacity for experiencing what others experience and identifying with others. 2 Butler, Five Sermons

66

certain experiences is if we already desired something else, and that desire was

satisfied. For instance, Butler writes “there could not be this pleasure, were it not for

that prior suitableness between the object and the passion: there could be no

enjoyment or delight from one thing more than another, from eating food more than

from swallowing a stone, if there were not an affection or appetite to one thing more

than another.”3 Other philosophers reason slightly differently, but the general

argument, which I call the ‘prior suitability’ argument, is the following: since

someone feels good after the satisfaction of a certain desire, say the desire for food,

this demonstrates that the desire existed prior to and independently of the desire to

feel good. The causal chain describing the pleasure we receive when we eat food

would look like the following4:

Desire for Food → Eating → Pleasure

If pleasure is something that occurs as a result of the desire for food, then the

pleasure did not cause the desire for food. Furthermore, it seems someone could

easily maintain that her desire for food is an intrinsically-held desire, something that

is desired for its own sake, in which case, such a person, at least when it comes to her

desire for food, would not have a hedonistic desire.

This all seems true enough, but note that prior suitability between a desire for

food and the pleasure gained from eating food does not refute hedonism. The prior

suitability line of reasoning does not tell us anything about the relationship between

the desire for food and the ultimate desire for pleasure. If the desire for food is

3 Butler 1726, p. 227 4 S&W, p. 278

67

caused by the ultimate desire for pleasure, then Butler’s reasoning is unconvincing as

a refutation of hedonism. So we need to know more. Butler’s argument raises the

following question, which I call the ‘prior cause’ question: what caused the desire for

food? If this explanation appeals to internal rewards, and for most people most of the

time it seems it would, then it would appear that hedonism, and hence egoism, is

vindicated right? Well, actually, there is another way to interpret the internal rewards

response, and perhaps even Butler’s prior suitability argument, that does not implicate

hedonism. What if the answer to the prior cause question does not cite a more

ultimate desire for internal rewards, or any desire at all? What if the answer cites a

feeling or emotion rather than a desire?

2.3 The Central Role of Feelings

The Butlerian type of argument that attempts to rule out hedonistic

explanations holds that the fact that someone feels pleasure after the satisfaction of a

certain desire, say the desire for food, demonstrates that the desire existed prior to the

desire to feel good. But the prior suitability argument is not sufficient to refute

hedonism because we will still want to know what caused the desire for food. If the

explanation refers to internal rewards, then it would appear that hedonism, and hence

egoism, is vindicated.

As I just suggested above, however, the answer to the prior-cause question

does not need to cite a more ultimate desire for internal rewards, or any desire at all.

What if the answer cites a feeling rather than a desire? In that case, it seems we do not

68

need to press further about more ultimate desires. Instead, it seems our inquiry at that

point ought to be one of understanding the nature of the underlying cause of the

feeling. There are three separate reasons for this, and more generally, for diverging

from the standard view of altruism underlying the obstacle to the prior suitability

argument. I turn to these three reasons next

2.3.1 Reason I

The first point is a point of language, but is not only a point of language. It is

also an observation that might lend support to the hypothesis that genuine altruism is

widespread. Since the question of altruism is not just about the existence of any

altruism (one instance of genuine altruism in the history of humanity would

presumably be insufficient to convince skeptics) but also about its prevalence, this

will turn out to be an important point. We frequently use the language of hedonism to

describe our motivations in very general terms, masking more specific explanations

that, on closer inspection, may turn out to be anything but egoism. Consider how we

often describe a reason for helping someone in terms of the feelings that are stirred by

seeing someone being harmed. As mentioned above, this fact gives egoists fodder for

believing that most people are not genuine altruists. But inspecting the full context of

a response like this, including queries for further explanations and subsequent

responses, may support the view that what lurks behind this response is genuine

altruism indeed.

69

To illustrate this point, I return to S&W’s example of a parent tending to her

child’s welfare. If I have a desire to take care of my child (an altruistic desire) and

one asks “why do you want to take care of your child?”, if there is no further desire to

which to refer then the desire to take care of my child is ultimate, and according to

S&W, it is the real deal—genuine altruism. On the other hand, if I give an

explanation that cites my feelings, such as “because it feels good to take care of my

child”, then the egoist could argue that this implies that I have the further, more

ultimate desire to feel good and so my desire to take care of my child is merely an

instrumental desire, one along the road to egoistic hedonism. However, if I were to

say that I feel bad when my child is harmed and this is the reason that I take action to

ensure his welfare, this is not enough for the proponent of egoism to win the debate in

terms of the standard definition of altruism. For the egoist to win, he must get me to

explain my feeling in terms of a prior desire. This point can become clear by

focusing on two different ways of probing further into my feelings:

Question 1: “why does it matter to you that it feel good to take care of your child?” Question 2: “why does it feel good to take care of your child?”

If one asks the first question then perhaps an appropriate response would cite

a more ultimate desire to feel good. Here, one would not be asking about the actual

cause of my feeling, but rather asking me to tell them why my feelings matter to me.

This way of probing might aptly call for the response to cite a desire, something with

propositional content.5 If instead the probing takes the form of Question 2, “why

5 To Question 1, “why does it matter to you that it feel good to take care of your children?” it is not clear that I would want to cite a desire here either. I could just as easily respond that the reason it

70

does it feel good to take care of your child?”, then it seems that what is being is asked

about the cause of my feelings—a more specific reason for why taking care of my

child feels good, and this more specific reason could be an emotion. As was

discussed in the above Appendix, many researchers maintain a distinction between

emotion and feeling. For instance, feelings of hunger would not qualify as an

emotion. Indeed, Prinz argues that emotions are a particular kind of feelings, whereas

feelings represent a more general category of affective states including hunger,

fatigue, emotions, tickles and sensations. When inquiring about the nature or cause of

a feeling of pleasure in taking care of my child, it seems that what is being requested

is a more specific type of feeling, and an emotion is the best candidate here. Feeling

hungry or tired or feeling an itch or a headache would hardly seem to constitute

reasons for why taking care of one’s child feels good. An adequate response to the

kind of inquiry reflected in Question 2 might be something like “I feel love, deep

affection, and sympathy for him and so it feels good to help him”—a response that

does not cite a prior desire.

Of course even if one could successfully determine that I also, rather

plausibly, simultaneously have the desire to feel good, this does not mean that my

desire to feel good is the cause of my desire to help my child. Admittedly, there is

also another way that a desire could be implicated in causing a feeling, and this is

when a feeling results from the satisfaction of the desire. Returning to the construal

of Butler’s example, one feels pleasure when a desire for food is satisfied. Likewise, I

matters to me is because I love them. The way of posing the question assumes that there is a desire to feel pleasure that is more ultimate, but it need not be.

71

could feel pleasure when my desire to take care of my child is satisfied. But of course

merely having the desire to take care of my child alone is not enough to account for

feelings of pleasure—my desire must also be satisfied. And when we are inquiring

into why it feels good to take care of my child, what we want to know is whether

there are any reasons prior to having the desire satisfied for why taking care of my

child feels good. It would not seem that my desire, by itself, is the reason for my

feelings of pleasure. More generally, it seems implausible that desires by themselves

cause feelings. To the question “why does it feel good to take care of your children?”

it would seem to greatly strain the meaning of the concept of desire to respond

“because I have a desire to feel good”. Whether this assertion is true or not, the

important point is that it is not necessary that we explain a feeling in terms of a prior

(egoistic) desire.

By focusing on alternative meanings behind a citation of internal rewards in

my first point I have somewhat shifted the debate from addressing the problem of

hedonistic desires. I have only shown that there is a way to cite feelings and avoid

the hedonistic trap. Again, I have done this because it seems important to the debate

on altruism to show not just that altruism exists, but that it is widespread. In our

explanations of altruistic behavior, we frequently refer to our feelings as the reasons,

and my intent was to show that this sort of referral alone is insufficient to indicate

egoism. The main problem with the Butlerian argument, however, was that if we cite

a more ultimate desire to avoid pain as our reason for helping someone, then prior

suitability fails to refute egoism. I will address this now by turning to my second, and

more important, reason for diverging from the standard view. My response here is

72

closely related to my first point: the argument against prior suitability, and the view

of altruism that underlies it, draws a sharp boundary between feelings and altruism,

but, as we will see, feelings are essential to genuine altruism.

73

2.3.2 Reason I I

Once again, the standard view maintains that genuine altruistic motivation

consists in an ultimate desire for the improving or protecting the welfare of another.

In order to understand what is missing from this view of altruism, it may help to once

again look at S&W’s demonstration of the contrast between altruistic ultimate desires

and egoistic ultimate desires. As before, the altruist (A) is represented in the first

column, and the egoist (E) in the second. But here, I have modified the diagram

somewhat in order to capture what was missing in S&W’s construal—namely, the

contrast between an ultimate altruistic desire and an instrumental altruistic desire.

Remember that the hard problem of genuine altruism is not in establishing that

we have desires to promote the welfare of others, for surely we do. This fact is what

makes egoism, and hedonism in particular, so tough to topple from its privileged

perch in much of the sciences and social sciences. The problem that proponents of

genuine altruism face is to distinguish the desires to help others that a genuine altruist

can be said to have from the desires to help others that an egoist can be said to have.

In the case of parental concern, A’s desire for her child’s welfare is ultimate. E’s

ultimate desire is to avoid pain, but this desire gives rise to her instrumental desire to

help her child. So both motivational pathways involve an altruistic desire. We now

need to determine whether it is obvious that one of them is more genuine than the

other.

In the diagram below, I have represented E’s desire for her child’s welfare as

an instrumental one, with her ultimate desire being that of minimizing pain.

74

Figure 3

Child needs help → Parent believes Child needs help

↓ ↓

A’s Desire for E Feels Bad that Child’s Welfare child needs help

↓ ↓ E’s Desire to

Minimize Pain ↓ ↓

A Helps Child E’s Desire for Child’s Welfare

↓ ↓ E Helps Child

If we focus on desires as the ultimate reason why each parent helps her child,

then E helps her child because she desires not to feel bad, whereas A helps her child

because she desires to promote her child’s welfare. But focusing on the contrast

between E and A in this way forces us to miss another important consideration—what

triggers the desire to help in each case? Remember from Chapter 1 that the D/I

Principle states that given two mechanisms for producing a behavior, the direct one

will be more reliable than the indirect one. Because E’s desire to help her child is

mediated through her feelings, there is an extra step involved when E helps her child

than when A helps her child. So an altruistic desire is a more direct mechanism for

helping behavior than is an egoistic desire. This is still the case even with the addition

of ultimate desires in the diagram above.

E’s desire to help her child is triggered by a feeling and not directly by a

belief. A’s desire to help her child is triggered directly by a belief. But what this

75

means is that while feelings are a necessary part of E’s motivational path, the same is

not true of A’s motivational path. This does not mean of course that feelings are not a

part of A’s altruistic motivation. More generally, this does not mean that altruistic

ultimate desires do not involve feelings in some way. As S&W argue, certain

feelings, namely sympathy and empathy, can sometimes and even frequently be the

cause of altruistic ultimate desires. However, they argue that while common-sense

suggests that empathy and sympathy trigger altruistic desires, it is still an open

question whether the ensuing desire is instrumental or ultimate. Perhaps, they

suggest, these feelings trigger altruistic desires only because people do not like

having these feelings and want to help others as a means of reducing these feelings.

So, according to them, when I learn that my friend is being harmed, I feel bad, and

feeling bad is an aversive feeling that I want to eliminate.

S&W recognize the distinction between feeling bad “for someone else”, a

sentiment typically associated with sympathy rather than empathy, and simply feeling

bad because I learn of some misfortune or suffering. They agree with much of the

relevant literature that the latter kind of feeling is more appropriately labeled personal

distress. As the name implies, personal distress is focused on the self and involves

feeling bad but without feeling bad for someone else. Thus, with the distinction

between sympathy and personal distress in place, it might seem as if there is a way to

classify a resulting desire as altruistic or egoistic—my desires triggered by personal

distress are acquired instrumentally merely so that I can stop feeling bad whereas my

desires triggered by sympathy are directly concerned with helping another person.

For the standard philosophical definition of altruism, however, there is obviously still

76

a problem with this response. Even if my feelings of sympathy trigger a desire to

help another, there could still be a hedonistic desire lurking in the background that

tells me to get rid of sympathetic feelings for others by any means possible. I will

return to say more about this below.

Aside from this problem, S&W argue that feelings of sympathy cannot come

to the rescue of altruism anyway. This is because they think that these feelings, and

others like them, are not a necessary part or cause of altruistic desires. They argue

along with other contemporary researchers that feelings and altruistic desires can

come apart. For instance, borrowing an example from Nancy Eisenberg, they write

“It is possible to enter these emotional states by thinking about problems that have

already been solved. Suppose Wendy discovers her husband’s infidelity, divorces

him, and then creates a good life for herself. If she then recounts this sequence of

events to Walter, he may find himself empathizing with the Wendy of a few years

past. What does this empathy motivate Walter to do? Apparently Walter empathizes

without forming the desire that Wendy’s situation be improved.” 6 Their point is

that feelings for and about another’s plight do not always cause a desire to help. It is

easy to see problems with the example S&W have chosen. Even though Walter does

not form a desire with the content “help Wendy divorce this guy” or “help Wendy

seek counseling to get over him” since these events have already occurred, there are

other kinds of desires that Walter could form. He might form a desire about the past,

of course, such as a desire that Wendy never had experienced these painful moments,

but this is probably closer to the concept of a wish than a desire since it would not be 6 Eisenberg, Nancy. 1986. Altruistic Emotion, Cognition, and Behavior. Hillsdale, N.J.: Lawrence Erlbaum Associates.

77

motivating in the same way. For this example, then, think of a desire that Walter

might form about the present, such as to ensure that Wendy really has moved on and

to express that he is always ready to listen. These are not trivial altruistic acts. They

consume time and energy. Despite being able to conceive of possible altruistic

desires that Walter forms in response to feeling sympathy, however, it does seem

appropriate to keep the concepts of emotions and desires distinct, as I have discussed

before and will say more about at the end of this chapter. It seems we can have

emotions of empathy and sympathy without forming an altruistic desire.

But S&W also think the opposite is possible. They argue that we frequently

have altruistic desires without any such feelings, such as when we learn of disasters

and cruelty in distant places and form ultimate desires that the people in those places

fare well. 7 In other words, they think genuine altruism frequently occurs without

any feelings whatsoever. This is because they believe that the distinction between

instrumental and ultimate desires is all we need. As they rightly note, the distinction

between ultimate and instrumental desires “is not a technical and esoteric idea but a

concept that is familiar from everyday life. Some things we want for their own sakes;

other things we want only because we think they are means to some more ultimate

end.”8 While this distinction seems prima facie convincing as the basis of the

difference between egoism and genuine altruism, as we will see, a deeper exploration

of the points at issue makes it less plausible that their difference lies solely, mainly, or

even at all, in the structure of desires. In so-called detached examples of altruism,

7 Sober and Wilson, pp. 237 8 S&W, p. 200. See pp. 200-222 for a discussion of some of the issues surrounding the distinction between ultimate and instrumental desires.

78

where we form desires about the welfare of people in distant places, it seems true

enough that we would not always feel sympathy or any one particular emotion. But

some feeling or disposition for a feeling, perhaps of worry or even anger, might

accompany or cause the desire to help. So just because sympathy in particular is not

experienced, this does not mean that some other feeling is not experienced. In cases

where there is no affective component or disposition at all, I am prepared to bite the

bullet and say that these desires simply do not represent genuine psychological

altruism.

To see why, return to the diagram above. In A’s motivational pathway,

feelings do not enter the picture at all, or at the very least, they need not. What

triggers A’s desire is a belief. In a similar manner, we could describe many instances

of desires that are triggered directly by beliefs alone. When I acquire the belief that it

is raining, this might give rise to my desire to wear a raincoat, or from the belief that I

need a pen if I want to write I might obtain the desire for a pen, or upon having the

belief that I am low on gas I desire to fill my gas tank, and so on. The important point

is that nowhere in any of these descriptions is any kind of feeling. Each desire is

keyed directly to a belief and not to any other mental state. In the case of staying dry,

of writing, and of driving, it might be perfectly accurate to describe the motivations

behind these actions as not involving, and never involving, feelings. But is it accurate

to describe genuine altruism in this way, as occurring in the absence of feelings,

without any role for feelings? I do not think it is. Altruism is different from these

other motivations, and not just in terms of representing different kinds of objects.

Altruism is a concept in everyday usage that is associated with feelings of sympathy

79

and love, or with anger and outrage. These feelings for others are what gives altruism

its force; without them altruism is merely a desire like any other, a representation of

some proposition that we want to be true, such as “water in the glass” if I want there

to be water in the glass.

Our everyday sense of altruism involves emotions, and without them altruism

is merely a cold representation of something we want to be true. Of course, even if

one agrees that certain emotions are a necessary part of genuine altruism, one could

maintain that ultimate desires are also a necessary part. According to this line of

argument, we would still face the task of determining whether a given altruistic desire

is instrumental or ultimate in order to know whether someone is an egoist or an

altruist. Once we establish that emotions are necessary for altruism, then, we will

want to know just how important their role is: are they the determining factor between

egoism and altruism or is the structure of desire, as the objection would maintain, still

the most important thing? To answer this question, we need to construct each

motivational pathway as a representation not only of the structure of desire but also of

the specific, ultimate emotional cause of the desire to take care of one’s child.

Specifically, we need to distinguish the role of an instrumental vs. an ultimate desire

simultaneously while distinguishing two different emotional causes.

Here we have S&W’s diagram of the two different motivational paths,

modified somewhat by leaving out the role of beliefs since they would merely

complicate the diagram without making any differences. Since we are trying to

determine which parent is the altruist here and it is not a given, I have labeled the two

parents X and Y instead of A and E.

80

Figure 4

Child Needs Help

↓ ↓ X’s amusement → with child

X’s Desire for child’s welfare

Y feels bad that child needs help

←Y’s love for child

↓ ↓ Y’s Desire to minimize

pain

↓ ↓ Y’s Desire for child’s

welfare

↓ X helps child Y helps child

Both X and Y are presented has having different emotional causes behind their

respective desires to take care of their children. In this scenario, Y feels love for her

child, which is of course the common emotion parents experience towards their

children. X, however, is amused by her child—her child is an interesting pastime. I

have purposively chosen two very different kinds of emotions in order to isolate the

role of each emotion from the structure of desires. Furthermore, because of the

difference in the structure of desire between X and Y, subsequent descriptions of the

role emotion plays in each will be different: because she is amused by her child, X

desires her child’s welfare; because she loves her child, Y feels bad when her child is

harmed. Even though emotions play a role in both parents’ desires, there is an

important difference between their desires for their children’s welfare. Y’s ultimate

desire is to avoid pain/seek pleasure, whereas X’s desire for her child’s welfare is

ultimate. The way that her emotion gives rise to her altruistic desire is similar to the

way S&W discussed the possibility of sympathetic emotions giving rise to other

altruistic desires—both then and now, the altruistic desire is ultimate.

81

Thus, in the diagram given here, the standard difference between an egoist and

an altruist is represented without blurring the line between the two. Now, which of the

two motivational pathways do we want to call more genuine than the other? Y’s

desire for her child’s welfare reduces to another desire, to avoid pain. This much is

still true. But Y’s desire for her child’s welfare also reduces, in an important sense, to

her love for her child. The cause of Y feeling bad when her child is harmed, and

hence the more ultimate cause of her desire for her child’s welfare is the love she

feels for her child. On the other hand, X’s desire for her child’s welfare does not

reduce to any other desire. However, it does, in an important sense, reduce to

something else—her amusement with her child. Between the two, I am more inclined

to say that Y is the more genuine altruist. The difference between X and Y can be

brought out even more if we treat Y’s desire to avoid pain as falling outside of the

causal path. This could be because Y’s desire to avoid pain is a background desire

that is not directly activated when she sees that her child needs help, or it could even

be, as S&W, think is perfectly possible, an unconscious desire. They write that “the

fact that individuals sincerely want to help others, and do not consciously experience

this desire as involving a sacrifice or a conflict with their authentic selves, does not

tell us what their ultimate motives really are. There is nothing in the egoism

hypothesis that prohibits other-directed desires from being fully integrated into the

agent’s personality. Those desires must be instrumental, but people need not

experience them as alien intrusions.”9 If Y’s desire to avoid pain is not experienced

but her love for her child is, and if what is experienced for X is her amusement with

9 S&W, p. 226

82

her child, then there seems to be an even larger difference in the genuineness between

X and Y’s altruistic motivation. Even if Y’s desire to avoid pain is not unconscious,

though, the point remains that in terms of the emotions felt, the more ultimate

motivation for her helping her child is her feeling of love, whereas the more ultimate

reason for X to help her child is her feeling of amusement.

An immediate objection to what I have just demonstrated might be that if X’s

desire for her child’s welfare is caused by her amusement with her child, then this

means that this desire would not be an ultimate one. Instead, a desire for amusement

would be ultimate—her feelings of amusement could be viewed as being constitutive

of a desire for amusement or a desire for amusement could get activated by feelings

of amusement. Either way, her desire for amusement would be more ultimate than,

and would cause, a desire for her child’s welfare. Her desire for her child’s welfare

then would be an instrumental one. This might be a plausible description under some

conceptions of emotions, such as those sometimes attributed to Hume, in which

feelings and desires are brought closer together. However, in many accounts of

emotions, such as the somatic feeling and behaviorist views discussed above, feelings

and desires are distinct. If feelings and desires are thought to be distinct, it is

possible that X’s amusement with her child would give rise to an ultimate desire for

her child’s welfare and not first, or at all, to a desire for amusement. As I discussed in

the previous chapter, it does not always seem to be the case that a given feeling gives

rise to a specific desire. The feeling of being angry at someone may give rise to a

desire to seek revenge or it may give rise to a desire to kick a shoe or even avoid that

person. Thus it is plausible to maintain that feelings of amusement would give rise to

83

X’s desire to take care of her child, without first forming some other desire. Of

course one might insist that feelings of amusement do not give rise to ultimate

altruistic desires, whereas feelings of sympathy and love always do. I hope to

demonstrate, however, that even granted this conceptual pairing between certain

feelings and a particular level of desire, it would still be the feelings that would be

doing the work in our understanding of genuine altruistic motivation and not the level

of the desire. Furthermore, it does seem plausible, as S&W have argued, that feelings

of sympathy, at least, would sometimes bring about only instrumental desires. Thus it

is not clear that this conceptual pairing should be granted.

If one adopts the perspective that desires and feelings are not distinct, but

rather two different components of the same mental state—one conative, or

motivating, and one affective—there are even better reasons for maintaining that

feelings play a more prominent role in our understanding of genuine altruism than

does the level of the desire. As discussed before, feelings for others are what gives

altruism its force; without them altruism is merely a desire like any other, a

representation of some proposition that we want to be true, such as “water in the

glass” if you want there to be water in the glass. Notice that even a regular old desire

like wanting water raises the question “why do you want water?” In other words, why

do you want the proposition “water in the glass” to be true? It is probably because

you are thirsty or care about your health. But even if I do not want to know more

about your desire for water, I would not doubt the authenticity of your wanting water.

The genuineness of your desire is not something that would be at issue. Of course, the

84

same is not true of altruism, since the central debate about altruism is about

distinguishing genuine altruism from apparent altruism.

The search for genuine altruism is an inquiry about fundamental motivations.

The point of trying to discover whether a particular altruistic desire is ultimate or

instrumental is to figure out what, at the core or sincerely, motivates someone to help

someone else. We want to know what they really care about and we will not be

satisfied until we have gotten to the root of the motivation. But why does this

motivation have to come in the form of desire? Furthermore why does inquiring into

what someone really cares about imply that only an ultimate motivation will suffice

in the response? In terms of the first question of why an altruistic motivation must

come in the form of a desire, we saw before that feelings can motivate. This is not to

suggest that feelings are always motivating, or that all feelings are motivating. All we

need is the more modest thesis that certain feelings are motivating at certain times, or,

even more minimally, that feelings give rise to desires and these desires are what

motivates. To be clear, I am not arguing that the idea of altruism would make more

sense if we only included affect apart from any conative factor, whether a desire or

something else. Even if one were to insist that feelings do not motivate on their own

and thus that they require, in addition, an altruistic desire, my concern is whether the

fact that a desire is ultimate is what determines whether an altruistic motivation is a

case of genuine altruism or whether, alternatively, it is the affective component that

determines this.

When we inquire into the cause of some altruistic action, there are at least two

different ways of posing the question. Thus far, the discussion of altruism has begun

85

with the question of ‘what does one ultimately desire or want?’ But there is another

way of posing the question of genuine altruism and it is to ask, again, ‘what does one

really care about?’ To see that they are different kinds of queries, consider an

example that is irrelevant to the altruism and egoism context. If you tell me that you

have an ultimate desire for doing some trivial activity, such as picking up a pen, not

for the purpose of writing anything, but only for its own sake, I would not think that

this is sufficient for indicating that you really care about the pen or about picking up

the pen. I would know that you desired to pick up the pen for its on sake, but whether

this meant that you really cared about the pen or not would seem to be an open

question. It seems the latter is our concern in the context of genuine altruism and the

pen example illustrates that it would not be satisfactorily resolved by citing an

ultimate desire. I will return to say more about the insufficiency of citing an ultimate

desire in the next section.

If the question of genuine altruism is ‘what does one really care about’, then a

response that cites a feeling would seem to satisfy the query. In the first place, caring

is often something that the subject does consciously, not something that is

unconscious or hidden to the subject. Frequently, we would not say that you care

about some object if you had no awareness of your feelings of concern for that object,

even if you take actions on behalf of it. Say that you work very hard and as a

consequence achieve great success in your career according to some material

standards. If you have no interest in being ambitious in this sense, and instead

sincerely deny ever being concerned about your material success, it would be odd to

insist that you really do care about your material success but you simply are not

86

aware of it. In fact, we say that people who do not ever consciously consider their

material success in their careers do not really care about being successful in that way.

We would say that other things motivate them, perhaps love for the work or perhaps

that they do not love the work but are just really good at what they do.

Of course in many cases it seems likely that people can come to discover that

they care about something by finding themselves acting to protect it, even if they

were previously not aware of any feelings of concern for it.10 However, despite

occasions where people lack awareness of feelings and yet still seem to care about the

object they are acting for, there is a more relevant point for the purposes here. Cases

where people do experience feelings of concern would also seem to be cases of

caring. In fact, the experience of feelings of concern for an object while taking actions

on behalf of it would seem to be sufficient for caring. There are a variety of examples

that would support this, but to take just one, consider someone working for

Greenpeace who feels great passion for the environmental work she does. In this case

and others like it, it seems obvious the person ‘really cares’ for the object in question.

It seems that really caring about something frequently requires, or at least

would be satisfied by, a conscious mental state—something with a phenomenological

feel—even if only from time to time, such as when one ponders the object in question

or reaffirms an interest in it after moments of doubt. But if this is the case, then the

best candidate for indicating whether someone cares about something is a feeling,

because a feeling just is something a subject is aware of. We saw earlier that S&W

and other philosophers maintain that we can have unconscious ultimate desires

10 Allen Buchanan made this very helpful point to me.

87

lurking in the background, but it is far less plausible that we could have unconscious

feelings. To take a very general example of a feeling, consider pain. We would not

insist that a person is in pain if she does not feel pain. It is the experience of pain

which defines pain. We would say that this person has suffered an injury but does not

feel it or at lest does not yet feel it. For an example of a feeling that is tied with

emotions, consider someone who loses her child. If she is in shock and sincerely

reports not feeling any sadness, then it would not make sense to insist that she does in

fact feel sadness but simply does not know it. It would be more appropriate to say

that she does not feel sad yet because she has not internalized or accepted what has

happened—that she is still in shock. While we can be mistaken about or unaware of

our ultimate desires even after sincere reflection and under normal circumstances, this

seems far less true when we consider our feelings precisely because they do have a

phenomenological feel. This is what they consist in. Thus, while a sincere avowal of

a feeling seems to be sufficient for understanding what one really cares about, a

sincere avowal of an ultimate desire does not seem to be.

2.3.3 Reason I I I

If our interest is in getting to the psychological heart of the matter, and we

want to know what someone really cares about, then it does not seem as if we will be

satisfied with knowing that one has an ultimate desire for another’s welfare. A critic

of what I have just argued might respond that the question of ‘what does one really

care about’ can be answered satisfactorily with the citation of an ultimate desire

88

because of what a desire being ultimate is meant to indicate about genuine

motivation. What we must do then is examine precisely what the appeal to ultimate

desires on the part of the theorist is supposed to accomplish and whether it succeeds

in this task. It seems there are three different criteria of genuine motivation that

ultimate desires are thought to satisfy. These are stability, strength, and uniqueness of

object. However, I will try to demonstrate that ultimate desires do not necessarily

satisfy these criteria for genuineness, giving us our third reason for rejecting the

standard view of altruism.

The main philosophical account of ultimate desires maintains that they are

more stable than instrumental ones. S&W elaborate this position most clearly by

arguing that the important difference between an instrumental desire and an ultimate

one is that when I desire something only as a means to some other end, I am more

likely to give up the means in light of new information. When I learn that my

instrumental desire no longer brings about the satisfaction of my ultimate one, I will

give it up. In addition, anything that causes me to give up my ultimate desire also

causes me to purge my instrumental desire, while the converse is not true: when I lose

an instrumental desire this says nothing about the ultimate desire that it served.

Theorists do not think that ultimate desires are impervious to change, only that

instrumental desires are much less impervious to change. Likewise, we can add that it

seems that an ultimate desire would be stronger for related reasons. One way in

which it would be stronger than the corresponding instrumental desire is if I were to

have reason to doubt the success of the instrumental desire in causally contributing to

the ultimate one. For even if I do not purge the instrumental desire from my set of

89

desires but only lose my faith in it, I may cease at any moment to be motivated by it,

or, at the very least, its motivational strength may decrease, while I would still be

strongly motivated by the ultimate one. Thus, ultimate desires can be said to be both

stronger and more stable than instrumental desires. But what this view assumes is

that the linguistic structure of desire is equivalent to psychological strength and

stability and the latter does not always seem true.

I could easily have some very stable instrumental desires. Consider having the

desire to eat food only because it promotes my ultimate desire to live or only because

it promotes my ultimate desire for pleasure. Even though it would be an instrumental

desire, because of the causal importance of food to the end of living or experiencing

pleasure, and since the latter two are significant ultimate desires, my instrumental

desire for food would be very stable. Conversely, I could also easily have very

fleeting ultimate desires. Imagine the desire to read political cartoons, not because of

something else it promotes but as an end in itself. I simply like reading political

cartoons.11 But if I learned of a new pastime that I liked better I could lose this

ultimate desire altogether. Now compare my ultimate desire to read political cartoons

to my instrumental desire to publish in philosophy. I do not try to publish journal

articles in philosophy for the sake of publishing itself, but because it promotes my

ultimate desire to engage in the discourse of philosophy and to be a professional

philosopher. Even still, it is a much more stable desire for me than my ultimate desire

11 Even though we might be able to say that one gets pleasure out of reading the cartoons, we should resist the temptation to recast this desire to be an instrumental one. If we do that, then we could do the same to ultimate altruistic desires and then there really is no philosophical debate worthy of our attention. In order to take the concerns in the literature seriously, we need to be able to treat certain desires as ultimate without the temptation to reduce them to other desires and so this is also true of the desire to read political cartoons.

90

for reading cartoons because what publishing philosophy is instrumental to, namely,

my desire to be a philosopher, is a very stable desire, and even more importantly

publishing philosophy is causally very important to being a professional philosopher.

This last comparison brings out an important point. How we view the stability

and the strength of various instrumental and ultimate desires depends on the mode of

comparison. Compared to the ultimate desire that an instrumental desire serves, the

instrumental desire will probably always be weaker: my desire to publish philosophy

is weaker than my desire to be a professional philosopher. Even this fact might not

obtain if an instrumental desire is linked to more than one ultimate desire, as with the

example of eating food which serves both an ultimate desire to live and one to

experience pleasure. Because it serves two ultimate desires, I could lose one of them

and still retain my instrumental desire for food, rendering it more stable than each of

the ultimate desires it serves. Even more telling, if we take an instrumental desire,

such as my desire to publish philosophy, and compare it not to the ultimate desire it

serves but to another ultimate desire, there is no reason to think that the ultimate

desire will be stronger or more stable: my instrumental desire to publish philosophy is

stronger than my ultimate desire to read political cartoons. I may sometimes

experience weakness of will when it comes time to taking actions on behalf of

publishing philosophy and would momentarily prefer to read cartoons, but my desire

for the latter would not be stronger than my desire to publish philosophy all things

considered.

Likewise when it comes to altruistic desires, it is easy to imagine that an

instrumental one is stronger than an ultimate one. Picture a creature slightly less

91

cognitively developed than humans. Since he has not faced all of the environmental

complexities and variations that humans have, he has never developed more than a

single ultimate end—to avoid pain and seek pleasure. This hedonic creature still

cares very much about the welfare of his child because this is related to his ultimate

hedonistic desire. In fact, given the limited environmental and social variability,

there are only a few other instrumental desires that this creature has, such as desires

about food, sex, physical pain, hierarchy in his social group, and attachments with

other conspecifics. Now consider that S&W and other philosophers are right in their

conception of genuine altruism. As a human being, all of my genuine altruistic

desires are ultimate ones, but because of the complexities of our environment and

culture, I have many such desires. Among my many ultimate altruistic desires, I have

a desire to help strangers find their way in my city. It seems obvious to consider that

the hedonic creature’s instrumental desire for the welfare of his child is much

stronger than my ultimate desire to help strangers with directions. If you are unhappy

with the comparison to another kind of creature, imagine that your only ultimate

desire is to avoid pain and seek pleasure. One of your main instrumental desires is to

take care of your friends—their harm brings you great pain. Compared to your

ultimate hedonistic desire, your desire for the welfare of your friends will be weaker.

But when compared with my ultimate desire to help strangers with directions, it is

much less plausible to think that your desire for the welfare of your friends is weaker.

To represent this point more generally:

92

Figure 5

(A) Ultimate Desire: (B) Ultimate Desire

Avoid Pain Another’s Welfare

↕

(a) Instrumental Desire:

Another’s Welfare

The double arrow indicates both that (A) gives rise to (a) and that (a) satisfies

(A). (a) would be weaker than and less stable than (A), but we cannot infer any

relationship about strength or stability between (B) and (a). 12 For all we know, (a) is

stronger than (B). If we had more information about these desires we might be able

to say that (B) is stronger than (a). But the point is that there is nothing in the

linguistic representation of these desires itself that tells us this. It might be the case

that in general, or on average, ultimate desires are stronger and more stable than

instrumental ones, although even this is far from obvious for the reasons given above.

Even if true, however, the tendency for ultimate desires to be stronger or more stable

does not tell us anything about a particular ultimate desire. If I were to show you that

people tend to spend more as their income increases, this would not allow me to say

anything about the spending habits of a particular individual. While we might want

to say that spending frequently depends on income, it would not allow us to define

12 As discussed above, even the first point can be questioned. If (a) is only instrumental to (A) without being related to any other ultimate desires, then it seems (a) would certainly be weaker than and less stable than (A). However, if (a) is an instrumental desire that serves additional ultimate desires, then it could be stronger and more stable than (A).

93

spending in terms of an income increase or even a qualified kind of spending, such as

‘serious spending’, as long as there are cases where the two factors, spending and

income, come apart. Insofar as the claim that ultimate desires are more genuine than

instrumental ones depends on claims about stability and strength, this is an empirical

claim. Since it is empirical and, as it turns out, not always true, it does not make

sense to make the claim the defining aspect of genuineness.

I argued that the structure of desire in itself tells us little about whether a

desire is stronger or more stable. But there is another way of thinking about

genuineness and this is to say that the object of an ultimate desire is special and

unique. The object of an ultimate desire is irreplaceable with other objects; for the

only thing that satisfies the desire is this object. On the other hand, because

instrumental desires are substitutable, so are the objects of these desires. This idea is

intuitive but the notion of irreplaceability is not limited to the objects of desires. It is

not the case that the only way to gain this criterion of genuineness is through

distinctions in the level of desires. We can gain it in two other ways. The first way is

generalizeable to all kinds of desires, including but not specific to altruistic ones,

because it involves the relationship between desires. Consider again my desire to

publish philosophy that serves my ultimate desire to be a professional philosopher.

Because the desire for publishing philosophy is instrumental, if I discover that writing

fiction would better serve my end, I would discard the desire to publish philosophy.

But, and this is the point, I doubt I would discover this. There is something about

publishing philosophy that makes it qualified as a way of bringing about my end of

engaging in the discourse of philosophy. Through rejections, reviews and acceptance,

94

trying to publish hones one’s skills at presenting philosophical ideas and reasoning

through arguments. And public accessibility to one’s philosophical work invites

discourse with others over time which in turn allows one to reflect further on the

ideas. In this way, publishing philosophy is unique. It is uniquely related to my end

of being a philosopher.13 Of course one could argue that my desire for reading

political cartoons is even more unique because of its status as being ultimate. While it

is true that the object of an ultimate desire is special and unique in the sense that it is

the only thing that satisfies the desire, this might shift the burden of establishing

uniqueness to the desire. How strong is this desire? Over what period of time is it

unique? How many such ultimate desires does one have? If one has countless

ultimate desires, would we still consider the object of any one of them to be unique or

special?

My desire to read political cartoons is a desire that I can easily imagine

discarding, however unique, if I discover some other pastime. In a way, its

uniqueness contributes to the possibility that I will lose interest in it. I might get sick

of it. While my desire for reading political cartoons can be characterized as very

unique for a limited time, my desire for publishing philosophy is somewhat less

unique for a much longer period of time (my entire career) and is stronger than the

former desire.14 But perhaps it would make sense to assume that people have very

13 I do not want to say that publishing philosophy is constitutive of being a philosopher, because I want to retain the distinction between instrumental and ultimate desires. But if the reader is not convinced with this distinction, she can substitute this example with my earlier one in which eating brings about pleasure. 14 To put the comparison in terms of romantic relationships, my desire for reading political cartoons is more like a short-lived infatuation, whereas my desire for publishing philosophy is more like a committed, happy marriage. I am happy to say that the latter feels more genuine.

95

few ultimate desires—not for things like pens but for objects such as one’s well-being

and one’s family. Thus the object of an ultimate desire is thought to be unique

because ultimate desires are rare. However, why should we assume that ultimate

desires are rare? The most likely reason would be that there is a psychological limit

to the number of objects one can consciously care about, and that an ultimate desire is

said to represent something one consciously cares about. This is not of course the

common way in which philosophers describe ultimate desires. More importantly,

even if they were described in this way, as discussed under Reason II, this would

seem to bring us back to making feelings the defining feature of caring, and not the

level of the desire.

My comments about gaining the quality of uniqueness through the

relationship that exists between a particular instrumental and ultimate desire apply to

motivations of all kinds. There is another way that we can gain this criterion of

genuineness which applies only to altruistic motivation. Our everyday intuitions on

altruism might suggest that an ultimate altruistic desire is required because it is a

desire about or for another person that does not reduce to any desire that is about or

for oneself. Since I have already discussed the insufficiency of desires alone

constituting our notion of genuine altruism, the important consideration for this way

of describing genuineness is the aspect of another person. Unlike the uniqueness

quality that can be said to exist in any context in which an instrumental desire

uniquely satisfies a particular ultimate desire, now I want to focus on uniqueness of a

particular sort—the distinction between oneself and another person. To represent

your uniqueness from me, my motivation to help you must be about you and not me.

96

While it is beyond the scope of this project to discuss the different range of feelings

that are associated with altruistic motivation, I wish to point out that we can achieve

this element of a motivation being about someone else and not the self through

differences in feelings. To take a prominent feeling identified with altruism, as I

discussed above there is an important difference between sympathy, which can be

described as feeling bad for someone else, and personal distress. Personal distress as

with sympathy involves feeling bad when I learn of or witness some misfortune or

suffering. But, as the name implies, personal distress is focused on the self and

involves feeling bad without feeling bad for someone else. On the other hand,

sympathy is about another person and not oneself. It is no coincidence that sympathy

has been closely associated with altruism both historically, through the work of

Hume, Smith, and Butler, and in our everyday lives.

As I described in my first reason for disagreeing with the standard view of

altruism, if one cites a feeling like sympathy instead of a desire as her motivation for

helping another, it does not make sense to inquire into the desire that caused the

emotion. There might be other kinds of causes we would want to know about of

course: did feelings for the other emerge from close contact over the years; are they a

result of being members of the same group or community; are they due to shared

similarities or simply because the other individual is a being capable of suffering, etc.

However, as argued above, it would greatly strain the concepts of ‘desire’ and

‘feeling’ to inquire into the desire that caused a feeling. But there is an asymmetry

here in terms of explaining altruistic motivation. We will frequently want to know

the emotional causes of altruistic desires. Altruistic actions are not trivial ones like

97

picking up a pen. They are by definition costly to the self. For this reason, we do not

just want to know what someone’s desires are, but, why they have the desires they do.

The altruistic parent from before, A, had a desire to help her child that was triggered

by a belief. But, without more, her altruism was merely a cold representation of

something she wanted to be true. We will want to know why A has a desire to help

her child—in other words, whether she really cares about her child. It seems we will

only be satisfied in our inquiry if A feels love and affection for her child, or more

generally, that A feels sympathy and compassion for suffering others. Otherwise it

would not represent what we common-sensically see as genuine altruism.

2.4 Conclusion

To conclude this chapter and the previous one, I have discussed how the

standard conception of genuine altruism is misguided because it excludes any

essential reference to feelings. I have also argued that the search for genuine altruism

is an inquiry into what someone genuinely cares about, and that we have options

other than determining whether a desire is instrumental or ultimate for understanding

the notion of genuineness. Ultimate desires for the improvement of another person’s

welfare were said to be more genuine because of what they are thought to indicate

about motivation—strength, stability, and a unique object. However, I have argued

that whether a desire is ultimate does not always tell us much about these criteria.

What this means is that an ultimate desire fails to satisfy those criteria which were

said to render it a satisfactory response to the question “what does one really care

98

about?” On the other hand, I have attempted to demonstrate that citing certain

feelings such as sympathy would satisfy this question because caring is something

that implies the experience of feelings.

The contemporary effort to distinguish altruism from the pursuit of internal

rewards has divorced genuine altruistic motivation from feelings, but feelings are

important to both: (A) what it means to genuinely care about something, in general;

and especially, (B) the notion of altruism, as understood both historically and

intuitively. Furthermore, ultimate desires do not seem to be necessary for genuine

altruism. To put all this together, my conclusion is that certain feelings are sufficient

to indicate a case of genuine altruism. As discussed in my introduction to the

previous chapter, my intent here has not been to establish the particular range of

feelings that are important for altruism. I have rather tried to demonstrate that

something important is missing from the standard view, and this is the role of

feelings. Of course, I have not specified the extent that feelings operate in a notion of

genuine altruistic motivation. Instead, the goal was to clear away the conceptual

baggage that has notoriously prevented both an acceptance in rational choice theory

that genuine altruism exists and efforts to theoretically and empirically explore the

reach and context of altruistic motivation—something I begin to do in the Appendix

following this chapter. Once we account for the prominent role of feelings in

understanding altruism, and at the same time recognize the limited, and perhaps

absent, role of ultimate desires, it can free us from the hedonistic trap mentioned

before.

99

2.5 Appendix 2: Applications to Rational Choice

In this final part of my discussion of altruism, I extend my analysis of the

standard view of altruism to discuss some specific applications to rational choice

theory. For the sake of continuity with the terminology of ‘preference’ rather than

‘desire’ used in the rational choice literature, in what remains I will refer to the

standard view as the ‘preference view’ of altruism. I will first explore the

implications of my proposed way of understanding genuine altruism to some

theoretical issues and then discuss how this view of genuine altruism might bare on

empirical efforts to test for the prevalence and conditions of altruism.

2.5.1 New Avenues for Theoretical Research

Most of rational choice theory includes the notion of ‘revealed preference’

which maintains that choices reveal what someone’s preferences are. A common

criticism of revealed preference theory is that the theory does not allow that

individuals’ choices and behavior will sometimes, perhaps frequently, be irrational or

fail to reflect their ‘true’ or reflective preferences. 15 A further problem with revealed

preference theory is that it does not allow us to explain behavior in terms of

preferences—to say that preferences cause behavior. But even if we keep the

concepts of preference and choice distinct, there is a further concern in the context of

altruism; as explained before, the preference view of motivation makes no essential

15 For a brief discussion of some of the concerns with RPT, see Hausman, Daniel and McPherson, Michael. 1996. Economic Analysis and Moral Philosophy. Cambridge: Cambridge University Press, pp.27-29

100

reference to the role of emotions and therefore underplays the significance of feelings

in altruism.16

Even though I have been arguing that the question of whether a desire or

preference is ultimate is misguided, and perhaps irrelevant, there is another kind of

reduction that should be avoided, and this is the collapsing of emotions to

preferences. When rational choice theorists and especially economists have discussed

emotions, including altruistic ones, they have typically viewed them in terms of costs

and benefits to be weighed against material costs and benefits. Emotions are pains

and pleasures to be minimized or maximized respectively. They have simply been

added to the utility functions as objects of maximization and subject to marginal

utility analysis. While the modeling of emotions in this way may be useful for some

purposes, it seriously masks the seemingly important differences between the

emotions and material interests. As Elster has powerfully argued, the passions, which

involve emotions, and material interests are not always fungible.17

Albert O. Hirschman explored the development of the idea that the pursuit of

interests can moderate destructive passions, such as glory.18 A society in which

people are motivated by material interests as opposed to passions would be more

predictable and stable. Cooperation would be facilitated because of common

16 I explore this issue much more extensively in my Dissertation 17 He provides an example of an agent deliberating between stealing an expensive book from the library. On the one hand, the individual has material interests in stealing, and on the other, he wishes to avoid guilt because it is an adverse feeling. If guilt were nothing but a cost than the agent should be willing to buy something like a guilt-erasing pill if it was cheaper than the cost of the book. But Elster writes “I submit that no person who is capable of being deterred by guilt would buy the pill. In fact, he would feel guilty about buying it.” (p. 303). A person willing to take the guilt-erasing pill would not need it. (1999. Alchemies of the Mind: Rationality and the Emotions. Cambridge: Cambridge University Press) 18 Hirschman, A.O. 1977. The Passions and the Interests: The Political Arguments for Capitalism before its Triumph. University Park, PA: Pennsylvania State University Press.

101

interests; people would have the incentive to work together to protect their interests.

If the passions could be reduced to material interests, then passions such as glory

could be assuaged and pacified by the acquisition of material goods. Hirschman’s

analysis implicitly assumes a sharper distinction between the passions and material

interests.

There are also other differences between material interests and emotions that

are veiled in attempts to model the emotions. As discussed in the Appendix to

Chapter 1, a range of emotion researchers, from William James and Karl Lange to

Antonio Damasio and Jesse Prinz, have identified the emotions as involving patterns

of bodily changes that, through their negative, positive, or mixed feelings, urge us to

act.19 Emotions have a phenomenological feel. Furthermore, I argued before that

they are often presented to us as a force or urge so strong that they overwhelm every

other concern or interest. We could not easily capture the latter difference by

modeling the emotion as lexically prior to other interests and concerns, however, for

at least two reasons. First, the overwhelmingness of the emotion may be very

temporary, and second, the emotion under such circumstances is not just a cost or

benefit in its own right but becomes a lens from which we evaluate the other costs

and benefits, the other interests and concerns. It colors, whether for the better or

worse, our assessment of what our interests are and what weight we assign to these

19 James, Williams. 1890. The Principles of Psychology. Henry Holt and Company; Damasio AR. 1994. Descartes' Error: Emotion, Reason and the Human Brain. New York: Grosset/Putnam; Damasio, Antonio. 1999. The Feeling of What Happens: Body and Emotion in the Making of Consciousness, Harcourt Brace & Company; Lange, C.G. 1922. “The Emotions”, in C.G. Lange and W. James (Eds.), Baltimore: Williams and Wilkins; Prinz, Jesse. 2004. Gut Reactions: A Perceptual Theory of Emotion. New York: Oxford University Press. While there are differences in these views, these differences are not important to the purposes in this paper.

102

interests. Hence, passions such as glory would seem to resist substitution with

material interests.20

If emotions that might appropriately be characterized as egoistic should be

viewed as separate categories from material interests, and thus not capable of being

captured in terms of preferences, then it is even more plausible to maintain this

distinction when considering altruism. When we focus the discussion of emotions to

just altruistic emotions, there is an additional and significant distinction between

emotions and preferences that is collapsed in formal models—this is the distinction

between the self and another. At least in the case of the status passions, for instance,

such as glory, vanity, and ambition, the emotions involved are about the self in some

important ways. But, as discussed before, emotions involved with altruism are about

others. Representing altruism in utility functions would collapse this further

important distinction between something that is about oneself and something that is

about another. The utility that someone would derive from helping others would be

substitutable with the utility gained from promoting one’s own self-interests. We

would not know why one gains utility from the promotion of someone else’s welfare

or consumption, and therefore we would not yet understand why there is a preference

to help another—is it because one cares about the other or is it because altruistic

motivation is a means of attaining political or social acclaim, or some other reasons? 20 Rousseau’s view seems to have been that the satisfaction of passions, even those that might be characterized as egoistic, is not interchangeable with the satisfaction of interests. The feelings associated with amour-propre, or relative status, for instance, are aimed at bringing harm to others, not at bringing benefit to the self, which implies that they are not reducible to interests. (Rousseau, J.J. 1964. "On the Origin and Foundations of Inequality Among Men". In R.D. Masters (Ed.), First and Second Discourses. New York: St. Martin's Press) For an excellent discussion of Rousseau on this point as well as Hobbes, see Grant, Ruth, “Passions and Interests Revisited: the Psychological Foundations of Economics and Politics”, unpublished manuscript.

103

For moral and political behavior at the very least, we should want a richer description

of our psychology, and there will be at least some circumstances when these

differences matter. Economists’ efforts to model altruism as externalities in utility

functions or as mere arguments in a cost-benefit calculus do not go far enough in

terms of distinguishing the self from another and are not capable of adequately

representing altruism as involving feelings.21

As discussed above, emotion researchers describe feelings as being central to

emotions. Feelings, through their either negative or positive valence, propel or

compel us to act. For instance, the feeling of fear would represent a dangerous

situation through registering changes in the body, but would also represent the danger

as something that is bad or negative, and serve as an inner reinforcer to withdraw

from the dangerous object, to eliminate the situation, often in a spontaneous fashion.

Because they urge us to want more or less of an object or situation, emotions sound a

21 Robert Frank has made the emotions central to an account of altruism, but his identifying the emotions with altruism is problematic. There are two ways to understand the role emotions are playing—either in terms of their consequences or in terms of being motivations. If we view them in the first way, then all of the emotions in his view, since he takes an individual selectionist framework, contribute to the long-term or ultimate material interest of the person—from this perspective all emotions are egoistic. If we view them in the second way, all emotions are motivations that compete with motivations to pursue short-term self-interest. But in viewing them this way, as competing with self-interest, emotions are all altruistic (or moral, which he takes to be synonymous with altruism). Frank could argue that there can be egoistic feelings, arising from rational calculation. Indeed he seems to hold this belief. In this case, emotions could be either altruistic or egoistic. However, in order for Frank’s account of emotions as commitment-solving devices to succeed, either emotions cannot be egoistic or else egoistic emotions must be sufficiently weak as to not really count in the same category as the altruistic emotions. This is because in the short-run, taking egoistic actions also has short-term material gains on its side—such as those that come from cheating a business partner. In order to counteract both (1) egoistic feelings and (2) material gains that come from taking a short-term egoistic action in the reward system that Frank identifies, altruistic feelings would have to be systematically very strong, so strong that it might not make sense in holding egoistic feelings and altruistic ones in the same category. If altruistic emotions were not systematically much stronger than egoistic ones, then these altruistic emotions could not hope to solve the commitment problems. (Frank, Robert H. 1998. Passions Within Reason: The Strategic Role of the Emotions. New York: W. W. Norton and Company, Inc.)

104

lot like preferences. Even though above I concluded, in agreement with many

researchers, that emotions are, among other things, action-tendencies, there seems to

be an important qualification to add. Emotions seem to incite or compel us to act, and

in this way can be classified as action-tendencies—as motivations; but emotions do

not seem to have aims or intentions. Emotions are about something and so it makes

sense to describe the objects of emotions, and they do compel and motivate us

towards action. But they do not necessarily have an aim and especially do not always

have the aim of being realized or satisfied in the way that preferences do. Another,

but slightly different, way to put this is in terms of the notion of ‘direction of fit’.

Preferences have a direction of fit that emotions do not possess; the former aim to

bring the world into fit with them.22 For instance, a preference to help someone has

the aim to bring about a state of affairs in which that person is helped. The emotion

of sympathy, on the other hand, does not aim at any state of the world, whether it is to

help someone or even to feel sympathetic.

So far in this section, I have been speaking of passions interchangeably with

emotions, but they are not synonymous. While it might be natural to view both the

emotions and the passions as picking out the same type of phenomena, the description

of the emotions that emotion-researchers provide us suggests that at least many of the

political passions are combinations of particular emotions with particular preferences.

Plato, for instance, characterized vanity, ambition, and glory as some of the passions

that drive political life, but they also aimed at certain consequences. These aimed, for

22 For a good discussion of the different views of desires/preferences, see: Smith, Michael. 1994. The Moral Problem. Oxford: Blackwell Publishing. pp.111-119

105

instance, at gaining recognition and victory.23 In the Rhetoric, Aristotle suggests that

emotions involve both affective feelings, such as distress and pleasure, and desires for

action. Anger, for example, may be a distressing desire for revenge. Other passions,

however, such as shame, seem to have more kinship with the contemporary way

emotions are conceived.24 The terminology is not relevant for my purposes, but again

it is important to maintain the distinction between feelings, which are a central aspect

of emotions, and preferences. Feelings may frequently give rise to preferences,

which in turn have an aim at some behavior or choice, but as I argued above, much of

the time one feels angry one might not acquire the preference for revenge. Contrary

to preferences, feelings do not seem to have specific goals. I have now added an

additional claim: feelings do not seem to have any goals or aims whatsoever. This

means that it does not make sense to say that feelings can be satisfied in the same way

that preferences can. This claim gains support from the fact that feelings of anger can

persist even after someone has achieved revenge or any number of behaviors.

Feelings are not exhausted by being ‘satisfied’, but preferences are.

Even though feelings do not necessarily have a particular goal, such as

revenge, I argued before that they are often commands to do something, for instance,

to curse or kick a shoe. This aspect to feelings is important. Although there is no

specific goal or any goal at all associated with a given feeling, they often compel us

to act in some way or another. Because feelings do not aim to be realized or satisfied,

but are commands to do something, they allow us to make better sense of behavior

23 Plato. 1968. The Republic. New York: Basic Books 24 Aristotle. 1932, 1960. The Rhetoric of Aristotle. New York: Appleton Century Crofts

106

that is expressive rather than consequential or instrumental. Expressive behavior is

behavior that does not always aim to achieve any outcome, at least at any given point

or in any given circumstance. For instance, someone who feels bad for others will

sometimes be compelled to just express his feeling bad, even if no one observes or

witnesses the expression. He may not aspire to affect outcomes on every occasion.

Expressing one’s feelings without being able to change things can cultivate these

feelings and foster behavior on other occasions that would be aimed at improving the

welfare of others. Thus, the expression of feelings certainly seems to have some

relationship with particular consequences, but expressive behavior does not seem to

stand in a deterministic relation with particular outcomes, and for this additional

reason, would not be subject to cost-benefit considerations.

Especially if we view preferences in terms of choice, as revealed preference

theory does, there is a gap between preferences and feelings. We could of course still

say that expressive behavior aims to achieve expression, and thus that someone might

have preferences to express oneself, but it may not be illuminating to do so. How

could we meaningfully determine the costs and benefits of expression when it is not

for the purpose of achieving any result? What standards would we employ for

measuring costs and benefits? Furthermore, as discussed above, there is sometimes

an overwhelmingness to feelings, so much so that they are not just, if at all, objects of

maximization, but they also shade our evaluation of other costs and benefits.

Feelings, by causing expressive behavior, thus seem to have many

applications to domains of life involving family, community and identity—in other

words, those areas where we would expect to see a greater prevalence of altruistic

107

behavior. Importantly, there is little or no temptation to free ride if what I am

compelled to do is express my sense of solidarity, outrage, or sympathy– nobody else

can do these things for me. To take just one topic, consider the application of

altruistic feelings to voting in large elections. Because the costs of voting to the voter

seem to far outweigh the benefits he receives, voting behavior has famously caused

problems for rational choice theory. To explain voting behavior, we could of course

ascribe altruistic preferences to voters. In these explanations, instead of just seeking

their own welfare, voters would have a preference for the welfare of others. But

merely assuming an altruistic preference would not be sufficient—the benefits, albeit

to others, would not outweigh the costs to the (altruistic) voter. It would still be the

case that the marginal impact of any particular voter’s participation, whether on

private or public interests, will be very low. A central way researchers have

addressed this potential problem is by describing the voter as having a preference to

express his sense of civic duty or concern.25 While this kind of approach is

promising, it fails to offer a satisfying explanation.26 These explanations, by

ascribing preferences for expressive behavior, maintain the position that voters

maximize some kind of expressive value—they are still aiming at a result. Voting

would still be subject to cost-benefit calculation, and would suffer from the problem

of meaningful standards mentioned before. Furthermore, there is no reason to believe

25 A prominent attempt is by Riker and Ordenshook. 1968. “A Theory of the Calculus of Voting”, American Political Science Review, 62, pp. 25-42. Also see Brennan, G. and Lomasky, L. 1993. Democracy and Decision. Cambridge: Cambridge Univ. Press. 26 For a seminal discussion of how rational choice theorists have frequently accommodated empirical anomalies by expanding the range of preferences to the point that theories become tautological, see Green and Shapiro (1994, Pathologies of Rational Choice Theory; a Critique of Applications in Political Science, Yale University Press: New Haven, especially pp 47-53 and 87-97).

108

that the calculus would come out in favor of expressing oneself, unless we simply

“stipulate” that the benefits of expression will be very high, enough to offset costly

voting.27 But, again, the important question at that point would be why the benefits of

expression are thought to be so high.

If we do not always view altruism in terms of preferences, however, then the

motivation for voting need not be entirely, if at all, consequential. A voter may

simply be compelled to express his sympathy or concern for others and since he is not

attempting to maximize his expression, it would not be subject to cost-benefit

considerations. Of course, this is only a potential explanation and would still

mandate empirical confirmation if it is to be taken seriously. This last point brings

me to some remarks about the possibility of testing for altruistic motivation.

2.5.2 New Avenues for Empirical Research

I have argued that the effort to distinguish altruism from the pursuit of internal

rewards has divorced altruism from feelings; feelings are an important part of

altruism that may illuminate some significant kinds of behavior. Furthermore, I have

argued that altruistic feelings cannot be collapsed into and captured by preferences.

An important advantage that potentially follows from this last point concerns the

viability of testing for altruistic motivations.

27 Green and Shapiro demonstrate that the Riker and Ordenshook attempt to explain voting by adding a preference for performing one’s civic duty just stipulates that the psychic rewards from voting are greater than the costs. Likewise, adding preferences for expressing altruistic concern would seem to be culpable of similar stipulation.

109

Although altruism is often researched and discussed across many disciplines,

a range of definitions and conceptions are employed and many studies designed to

test for altruism fail to make important distinctions. For instance, most research in

economics, political science, and anthropology does not distinguish between altruism

as a behavior and altruism as a motivation; donations to charity, even when made

anonymously, for example, fail to satisfactorily demonstrate altruistic motivations.

Evolutionary game-theory is an increasingly useful field for understanding moral and

political behavior, including altruism, and compliance to social norms. But

contributions to this field, such as those by Kenneth Binmore, Robert Axelrod, and

Brian Skyrms, have only successfully modeled cooperation, fairness, or altruistic

punishment. 28 They have not been able to establish the existence of altruism—they

are only “how possible” and not “why necessary” explanations.29

The existing empirical research on altruism can be said to fall into two main

camps, neither of which adequately addresses the concerns about establishing

altruism. In the first camp there are the experimental studies, such as those in game

theory and behavioral psychology. Experimental or behavioral game-theory in

economics, political science, anthropology, and psychology has attempted to capture

emotional rather than rational motivations, but there are two major shortcomings to

this research. First, game-theory experiments have mainly investigated how fair or

trusting participants are in ‘dictator’ or ‘ultimatum’ games.30 Fairness and trust are

28 Binmore K. 1994. Playing Fair: Game Theory and the Social Contract I. MIT Press, Cambridge, MA; Axelrod R. 1984. The Evolution of Cooperation. Basic Books, New York 29 See Alexander Rosenberg for a good discussion of the difference. 30 The Dictator game examines how a subject will divide a resource between himself and another subject in a one-shot context. In the Ultimatum game, one person (proposer) is given the choice of

110

important ethical notions that may involve the emotions, but they are not identical to

altruism.31 In addition, there is some evidence suggesting that subjects respond to the

unconscious cues of testing environments and instructions. For instance, when

subjects are instructed to “divide” a resource, this may bias a subject towards sharing

a resource equally or fairly since “divide” seems to suggest an equal split.32 Thus

some of the ‘fairness’ results can be called into question.

The next, and perhaps even more important, problem with experimental game

studies is that they mainly focus on behavior and thus offer us thin descriptions of

actions, whereby individuals are said to act in accordance with certain theories or act

in a manner consistent with having particular motivations.33 For instance, the seminal

work by Eckel & Grossman investigates contexts in which subjects make donations to

others, but these studies do not give us thick descriptions of actions—i.e., those that

dividing a resource between herself and the second person (responder), who then has one chance to either accept the terms of the division as they have been given to her or else refuse, in which case she receives nothing at all. If the total resource is, say, $10, and only divisible in dollar units, then if the proposer is “fair”, she will offer $5 to the responder and keep $5 for herself. Standard rational choice theory, however, predicts that the proposer will keep $9 for herself and offer only $1 to the responder, because if the responder is rational she would prefer $1 to zero. Since the proposer knows that the responder is rational and is also rational herself, then she will divide the total to keep most for herself. 31 Many researchers view the fairness outcome (50-50 split) as altruistic. The belief is that when a proposer divides a resource fairly, this division is altruistic because he could have kept all of the resource for himself instead—thus even though the proposer does not divide the resource so that the responder gets more than she does, a 50-50 split is still worse than a proposer could have done in an Ultimatum game, as predicted by rational choice theory. The problem with seeing a 50-50 split as altruistic is that there is good reason to predict that the responder will reject an offer if it is not fair. People get angry at individuals who have “poor manners” (Camerer & Thaler, 1995). Thus the proposer might just be pragmatically trying to avoid a no-win situation. Supporting this argument is the fact that in the “Dictator” games, where only the proposer’s decision determines the division of a resource and where participants are anonymous, by far the most common result is for the proposer to offer nothing to the responder (Burnham, 2003). So this seems to strongly suggest that it is a concern with the threat of rejection, rather than altruism or even fairness, that is leading the proposer to offer 50-50 splits in many ultimatum experiments. 32 Smith, V. (2003). Experimental Methods in Economics. Encyclopedia of Cognitive Science, Lynn Nadel (ed.), New York: Nature Publishing Group, Macmillan Publishing, 1070-1079. 33 For results and discussions see especially Guth & Tietz, 1990; Forsythe et. al., 1994; Hoffman et. al., 1994, 1996; Eckel & Grossman, 1996; Burnham, 2003; and Henrich et. al., 2005.

111

would actually investigate individuals’ motivations.34 The main exception to this

work is that of Ernst Fehr and his colleagues who have attempted to find the neural

correlates involved with punishing free-riders when punishers incur a monetary

cost.35 However, even studies by Fehr and his colleagues on so-called ‘altruistic

punishment’ have examined a very restrictive sense of altruism, if one even calls it

altruism—altruistic punishment in their study could be interpreted to involve spite

which does not seem to be the same thing as altruism.

In the second camp, there are psychological experiments, especially by Batson

and Shaw, designed specifically for testing motivational altruism in contexts in which

subjects are in apparent distress. 36 Their valuable work shows us that it is

implausible to believe that very crude theories of self-interest, such as the pursuit of

material gain, can explain behavior in many circumstances. But critics argue that

these existing studies on altruism do not adequately rule out more sophisticated forms

of egoism—reductive altruism—and therefore do not adequately test for altruism.

This is because (a) apparently altruistic motivations are compatible with more

ultimate self-interested explanations (e.g., seeking internal rewards), and (b) even

subjects’ self-reports of their altruistic preferences cannot rule out unconscious self-

34 Eckel, C.C. and Grossman, P.J. (1996). Altruism in anonymous dictator games. Games and Economic Behavior, 16 (2), 181-191 35 Fehr, Ernst and Gächter, Simon. (2002). Altruistic Punishment in Humans, NATURE 415, 10, 137-140; Fehr, Ernst and Fischbacher, Urs. (2003). The Nature of Human Altruism, NATURE 425, 23, 785-791; Fehr, Ernst and Fischbacher, Urs. (2004). Social norms and human cooperation, TRENDS in Cognitive Sciences Vol.8 No.4, 185-190; Fehr, Ernst. (2004). The Neural Basis of Altruistic Punishment Science 305, 27, 1254-1258 36 Batson, C.D. (1992). Experimental tests for the existence of altruism. Proceedings of the Biennial Meeting of the Philosophy of Science Association, 2, 69-78; Batson, C.D. (1991). The Altruism Question: Toward a social-psychological answer. Hillsdale, NJ: Erlbaum; Batson, C.D., and Shaw, L.L. (1991). Encouraging words concerning the evidence for altruism. Psychological Inquiry, 2 (2), 159-168; Batson, C.D. and Shaw, L.L. (1991). Evidence for altruism: Toward a pluralism of prosocial motives.Psychological Inquiry, 2 (2), 107-122.

112

interest. More importantly, many researchers believe that we cannot determine

peoples’ ultimate preferences and goals. Therefore, they believe that there is no

objective, scientific method of studying altruistic motivation and that we will forever

be forced to rely on self-reports that can be misleading, erroneous, or reduced to self-

interested explanations. 37

Researchers may be correct in their pessimistic assessment that there are

significant obstacles to researching motivational altruism. But, importantly, they are

only correct if one accepts the standard contemporary view of genuine altruism as

being a function of ultimate preferences. It does seem plausible to argue that people’s

ultimate preferences are not something that we have adequate scientific access to.

Recall that researchers maintain that ultimate preferences can be unconsciously

held—the individual herself may not be aware of them.38 Furthermore, it is unlikely

that we can discover the psycho-physiological or neural correlates to preferences,

aside from whatever emotion or feeling component might occur—but then feelings or

emotions become central, and this is not the standard way of viewing preferences. If

we considered preferences as necessarily being affective, as involving feelings, as I

argued in the first Appendix, many of the ways in which we explain or make sense of

someone’s actions would not qualify as preferences. For instance, I currently want to

37 For discussions about the theoretical and empirical difficulties for testing altruism see: Zahn-Waxler, C. 1991. The case for empathy: A developmental perspective. Psychological Inquiry, 2 (2), 155-158; Hornstein, H.A. 1991. Empathic distress and altruism: Still inseparable. Psychological Inquiry, 2 (2), 133-135; Dovidio, J.F. 1991. The empathy-altruism hypothesis: Paradigm and promise. Psychological Inquiry, 2 (2), 126-128; Sober, E. (1991). The logic of the empathy-altruism hypothesis. Psychological Inquiry, 2 (2), 144-147; and Dennett, 2003. 38 A great deal has been written about the nature of desire in Hume. Michael Smith interprets Hume as maintaining that only some desires are experienced as feelings. More broadly, even aside from his reading of Hume, Smith makes a compelling case for the implausibility of viewing all desires as having a phenomenological feel. Smith, pp.92-10.1

113

continue typing, but it would be misleading to think of this preference in terms of an

emotion or as being triggered by an affective state.39 This consideration also explains

why first-person subjective reports are not viewed as adequate measures of peoples’

actual preferences. Since preferences are not necessarily affective, they do not

necessarily all have a phenomenological feel. Thus it is reasonable to hold that people

are not always aware of their preferences. Even if one were to hold that many

preferences are affective states or involve affective states, however, the important

point is that for the standard view of genuine altruism, altruistic preferences are not

essentially about affective states.

A significant advantage of the feeling view of genuine altruism over the

standard definition is that it may allow for greater testing for the conditions and

existence of genuine altruism. For instance, work distinguishing the neural and

physical correlates of emotions is quickly emerging.40 There is extensive work

supporting different physiological response patterns between empathy and sympathy

39 Also, consider a long-term preference, for instance, the long-term preference to be a doctor. In moments of reaffirmation of this preference, following periods of doubt, the preference to practice medicine might express itself with a kind of force and perhaps even a feeling. But even in the absence of this feeling, during times in which the desire is not being challenged or reassessed, it would be odd to say that this preference simply goes away. It would also be odd to say that people no longer have preferences or cease to have them when they are sleeping. 40 For instance, recent work published in NeuroImage by Jack Nitschke and his colleagues, 2003, suggests that the feelings of love that mothers experience with their babies is reflected in very strong activity in the orbitofrontal cortex. Also see: Davidson, R.J. & Irwin, W. 1999. The functional neuroanatomy of emotion and affective style. Trends in Cognitive Science, 3,11-21; Adolphs R, Tranel D, Damasio H, Damasio AR. 1995. Fear and the human amygdala. Journal of Neuroscience, 15(9):5879-91; See also the separate works on the role of emotional salience by: Haidt, J. 2001. The emotional dog and its rational tail: A social intuitionist approach to moral judgment. Psychological Review, 108, 814-834; Kensinger, E.A., and Corkin, S. 2004. Two routes to emotional memory: Distinct neural processes for valence and arousal. Proceedings of the National Academy of Sciences, USA, 101, 3310-3315; and, Dolcos, F., LaBar, K. S., & Cabeza, R.. 2004. Interaction between the amygdale and the medial temporal lobe memory system predicts better memory for emotional events. Neuron, 42, 855-863

114

on the one hand and personal distress on the other.41 To take just one difference,

sympathy is robustly associated with lowered heart rate, and personal distress with

increased heart rate. Furthermore, as discussed before, questions about motives are

difficult to gauge through self-reports. But the feelings-view would seem to privilege

first-personal reports.42 This is because while someone could plausibly have

unconscious ultimate desires lurking in the background, it is far less plausible that he

has unconscious feelings. While researchers maintain that we can be mistaken about

or unaware of our ultimate desires even after sincere reflection and under normal

circumstances, this seems far less true when we consider our feelings precisely

because, as argued before, they do have a phenomenological feel. Thus, the feelings

view of genuine altruism would seem to offer a potentially important avenue of

testing for altruism.

41 Eisenberg, N., Guthrie, I.K., Murphy, B.C., Shepard, S.A., Cumberland, A., and Carlo, G. (1999). Consistency and development of prosocial dispositions: A longitudinal study. Child Development, 70 (6), 1360-1372; Eisenberg, N., Losoya, S., & Spinrad, T. (2003). Affect and prosocial responding. In: R. J. Davidson, K. R. Scherer, & H. H. Goldsmith (Eds.). Handbook of Affective Sciences, New York: Oxford Univ. Press. 787-803. 42 One final difference between the two views of altruism is the following: whereas the preference-view maintains that there is one and only one kind of genuine altruism—the individual must have an ultimate desire for others’ welfare—the feelings-view is consistent with the recognition that there may be different feelings that are relevant to altruism depending upon the culture. After examining various cross-cultural research, Paul Eckman has famously argued that there are primary or basic feelings, such as anger and sadness, that are universal, and more sophisticated feelings, which are rooted in the basic ones, that vary across cultures. It may be the case that the feelings involved with altruism are of the latter sort. If this is true, then in order to study altruism in political life, we would first need to have an understanding of which feelings are central to a particular culture and then inquire into the prevalence of those feelings and the contexts in which they arise. (For instance, see Eckman, Paul. 1997. What the Face Reveals: Basic and Applied Studies of Spontaneous Expression Using the Facial Action Coding System. New York, NY: Oxford University Press)

115

Part I I Rational Choice and Consequential ism

3 The Structure of Moral Decision-making: Appreciating All of our Particular Ends

3.1 Introduction

This part of the dissertation turns to moral theory to focus on another feature

of rational choice theory: its emphasis on consequences as the locus for determining

whether a particular action, policy or institution is good or bad. As discussed in the

Introduction to the dissertation, there are two ways in which consequences feature in

rational choice theory. First, consequences, and in particular wellbeing measured by

some form of utility or preference-satisfaction, explicitly figure in welfare economics.

Welfare economics is the branch of rational choice theory dedicated to understanding

what economic policies or actions make individuals better off. Importantly, whether

or not someone is better off is determined by the relevant consequences. Second, and

more subtly, consequences figure in the approach of rational choice. The rational

choice theorist believes that to ask whether an action or policy is a good one is

unhelpful unless we know what the alternative to that action or policy is. Rational

choice is the process of choosing the best, however best is to be defined, selection

among available options. Furthermore, the rational choice theorist believes that to

ask whether one action or policy is better than another is already to imply a common

measurement from which comparisons can be made. By being comparative in

analysis, rational choice then places the emphasis on consequences, and specifically,

on general consequences, because they serve as the basis from which to make

116

comparisons between competing options. In both of these ways, rational choice

theory is a form of consequentialism and is therefore subject to the sorts of objections

typically raised against consequentialism as a moral theory.

Many contemporary moral philosophers oppose consequentialism because, it

is argued, it privileges abstract or general values over particular ones. For instance,

the only ends that are valued for their own sake in consequentialist theories are the

relatively general ones such as happiness and welfare. However, it is widely believed

that, contrary to consequentialism, it is particular ends such as honesty and family and

not general ones that are to be valued for their own sake.1 Related to this position on

the proper source of value is the view that consequentialism is a psychologically

demanding theory that alienates us from our own personal relationships and plans.2

More recently, and in a context largely unrelated to the problems of consequentialism,

Michael Smith argues that the morally good person is someone who is motivated

directly by her particular moral ends even when they are impartial or impersonal

ends.3 It is this last criticism that I will focus on in this chapter.

1For discussions of these and related criticisms, see: Philippa Foot’s “Utilitarianism and the Virtues,” and Peter Railton’s “Alienation, Consequentialism, and the Demands of Morality,” both in Consequentialism and its Critics, ed., by Samuel Scheffler. (New York: Oxford University Press, 1988); Owen Flanagan, in Varieties of Moral Personality: Ethics and Psychological Realism. Cambridge, Mass.: Harvard University Press, 1991, especially Chapters 3 and 4; Michael Stocker, in “The Schizophrenia of Modern Ethical Theories,” Journal of Philosophy, 73 (1976), 453-66; Bernard Williams, in Moral Luck: Philosophical Papers, 1973-1980. Cambridge: Cambridge University Press, 1981, especially “Persons, character, and morality.”; and, Lawrence Blum, “Against Deriving Particularity,” in Moral Particularism, ed., by Brad Hooker and Margaret Little. Oxford: Oxford University Press, 2000 2 The second reason concerns a supposed disharmony between one’s motivations for acting, and one’s justifications for acting. The only justification for acting under consequentialism is to maximize good consequences, however defined. But this reason for acting is a very demanding one and might be very different from the less general concerns a person actually has and from the particular ends and virtues that a morally good person ought to be motivated by. The result of such a psychological and moral disharmony can be a diminished moral life 3Smith, Michael. The Moral Problem. (Massachusetts: Blackwell Publishing), 1994.

117

There have been several attempts among consequentialist theorists to create

what I call ‘two-level’ theories in order to address such concerns. In two-level

theories there is a significant wedge between a justificatory general level of value, on

the one hand, and a motivational particular level of ends, on the other. While there is

a range of these sorts of theories and not all of them can be characterized as having

different ‘levels’, I use this term only to convey the central point that general values

and concerns are held to be independent of and take priority over particular ones in

important ways. Defenses of such two-level theories, however, do not succeed in

being full responses to the kinds of charges discussed above. While they seem to

successfully address the problem of demandingness and the problem of psychological

tensions, even the most sophisticated defense of such a theory—Peter Railton’s—

does not yet adequately or fully deal with the concerns about (1) the proper source of

our values; and, (2) the characterization of the good person. I address mainly (2)

here, although I briefly discuss how (1) and (2) are related concerns and thus require a

similar response.

After exploring the appropriate description of the good person, however, I

argue that not only are two-level theories, and in particular, the motivational

structures they contain, consistent with the appropriate view of the morally good

person, but they are critical to it. This is because they are important for trying to

reflect on our particular ends and for fully understanding and responding to the

significance of all of our particular ends when we have more than one. Rather than a

defense of consequentialism, and hence rational choice, based on its purported

coherence or rationality, in this chapter I offer a defense of the kind of

118

consequentialism embraced by Railton based on what it means to care for one’s moral

ends. Whereas Railton seems to argue that two-level consequentialists can be

committed to their ends despite their consequentialism, I hope to show that it is

because of the structure contained in these forms of consequentialism that

consequentialists can be committed to their particular ends.

3.2 Against General Value

In simple consequentialist theories, the only things that are said to be

intrinsically valuable—that is, valued for their own sake and not for the sake of other

things—are relatively general things such as happiness, welfare, preference-

satisfaction, and the like. However many scholars, especially in the virtue and care

ethics traditions, argue that it is those things which are not so general or abstract that

are intrinsically valuable. It is particular virtues and character traits, such as honesty,

benevolence, loyalty and treating people with respect and equality, and our partial

ends such as our families, relationships, and life projects that are intrinsically

valuable.4 Bernard Williams, for instance, argues that we are alienated from our deep

4 While it is possible that particularity and partiality do not amount to the same thing—one could hold that intrinsic value lies in specific yet non-partial virtues such as treating humanity with respect and dignity(Philippa Foot seems to maintain this possibility in “Utilitarianism and the Virtues”)—the two claims often go hand in hand. The proper objects of particular virtues such as honesty and loyalty are often those with whom we have special relationships, and indeed, with some virtues such as integrity, the object is oneself and perhaps nothing is more partial. Thus I will not distinguish here between literature that emphasize the value in particular virtues and literature that is concerned with the value in partiality. Instead I will lump both particularity and partiality under the heading of “particular” except for when the differences are important.( By ‘particular’ I am not referring to the way particular is used in the debate between moral theory and moral particularism) . But because I will confine my discussion to arguments about ends, I will not include agent-relativity of the sort discussed by Thomas Nagel where reasons for acting in particular sorts of ways towards others can be said to be relative to particular individuals.( See ‘Autonomy and Deontology’ in Consequentialism and its Critics.)

119

projects and loved ones when we are motivated by general moral ends.5 Williams’

example of a man forced to choose between saving the life of his wife or a stranger

illuminates his argument. If the man is moved by general and impartial values, then

his motivating thought to save his wife might be something like “it is my wife, and in

situations like this it’s permissible to save one’s wife”.6 But Williams famously

objects that this man has “one thought too many”. Instead, the man’s motivation to

save his wife should be out of direct concern and love for his wife; because she is his

wife, period. Likewise, Susan Wolf argues that theories such as utilitarianism force

us to strive for a kind of moral perfection.7 According to Wolf, utilitarianism is too

demanding on us given that our normal psychology includes being motivated by

desires and projects that are sometimes in tension with morality. More to the point,

however, even if moral perfection were not so demanding—consider one who derives

happiness from always helping and being kind to others—someone who attained

moral perfection would be someone we consider stilted, humorless, and one-

dimensional. We would not think it good to encourage anyone to be this way.

Wolf’s argument seems to apply more broadly to the valuing of general ends when

they are thought to be separate from particular ones. For it seems that the generalness

or abstractness of the end of happiness in utilitarianism is in large part what

contributes to it being a demanding theory. A view that would specify particular

commitments to family and kindness only for those nearest and dearest would

necessarily be more restrictive. Although this is not necessarily the case—consider a 5Williams, Bernard, “A Critique of Utilitarianism,” in J. J. C. Smart and Bernard Williams, Utilitarianism: For and Against. Cambridge: Cambridge University Press, 1973. 6 Ibid., p. 101 7 Wolf, Susan. “Moral Saints,” The Journal of Philosophy, Vol. LXXIX, No.8, August, 1982.

120

consequentialist theory that specifies only one fairly undemanding obligation such as

giving away $10 a year—the generalness of consequentialist ends seems to go hand

in hand with its demandingness.

To address these challenges about general values, some adherents of

consequentialism have introduced complex forms of the theory which contain ‘two-

level’ structures, in which there is an ultimate level of value and a lived level of ends

and traits. The ultimate level consists of some relatively general consequentialist

values such as total welfare or happiness, and the lived level consists of particular

practices, character traits, and ends that receive their justification through their

contribution to the good consequences. While there is a range of these sorts of

theories and not all of them can be characterized as having different ‘levels’, I use this

term only to convey the central point that general values are held to be independent of

and take priority over particular ends in important ways, as will be shown below. In

addition, it is possible to have more than two levels of values and ends, but this would

not affect the main issues in this chapter.

One two-level theory notable for its appreciation of the psychological

importance of our particular ends is Peter Railton’s.8 Railton argues that since simple

forms of consequentialism only regard good consequences, however defined, as

intrinsically valuable, they fail to take seriously other, more particular ends and this

detaches people from meaningful aspects of life. He proposes ‘sophisticated

consequentialism’ in which an agent acknowledges the value of promoting good

outcomes but at the same time acts by cultivating certain habits and character traits

8 Railton, “Alienation, Consequentialism, and the Demands of Morality”.

121

that enable particular commitments.9 A sophisticated consequentialist need not

always be consciously motivated by the end of good consequences and so may do

various things for their own sake. Because he does not need to bring consequentialist

deliberation to bear on every action, he consciously acts for the sake of his particular

commitments and so he does not suffer the fate of alienation.10 Importantly, Railton

demonstrates how his sophisticated consequentialism avoids collapsing into simple

act-consequentialism.11

It does seem that Railton’s two-level view would reduce the psychological

tensions problematic to simple consequentialist theories. If I am not always required

to conscientiously deliberate on good consequences in order to bring them about, I

can focus my attention on my family, friends, and the projects I hold dear. This will

be less alienating—in my every day life I will be consciously motivated by my

particular commitments and will thus be able to achieve a greater sense of emotional

fulfillment through my engagement with them. This will also be less demanding in

terms of my time and resources—I will focus on cultivating character traits and habits

that motivate me to attend to those with whom I have special relationships. I will not

need to constantly spend my time striving to help everyone in the world. Two-level

views, with their mandates to focus on particular ends, do seem to be a great

improvement over simple forms of consequentialism in terms of enabling an agent to

live a rich and varied life.

9 Railton, p. 109 10 Ibid., see especially p. 113 for a full explanation 11 Railton argues that unlike rule- and trait- consequentialism, sophisticated consequentialism is based on specific actions. A sophisticated consequentialist acts so as to maximize the good, but does not bring consequentialist deliberation to bear on each act. Because it is based on acts, he argues, sophisticated consequentialism does not collapse into standard act-consequentialism. pp.113-119.

122

However, two-level views come up against another sort of problem raised

most recently by Michael Smith that would seem to be damaging to consequentialism

as a moral theory—that they do not contain the proper description of the morally

good person. While William’s concern, mentioned above, applies to the treatment of

our personal ends, we will see that Smith’s concern is with the proper treatment of

even impartial or impersonal ends and is therefore potentially more damaging to

consequentialism. For even if one took a strong Williams-like point of view and

insisted that only impersonal and impartial ends fall within the domain of moral

theory, Smith’s concern would suggest that the consequentialist would still lack the

appropriate commitment to his particular (moral) ends.12

Smith begins with the claim that the good and strong-willed person (hereafter,

the good person) is someone who is reliably concerned for or motivated by his

particular value judgments. Smith’s goal is to argue against motive-externalism,

which maintains that one can judge a course of action to be right and not be motivated

to take that action.13 However his analysis also presents an apparent problem for

consequentialism, even in its two-level form. In arguing against views that maintain

a distinction between judgment and motivation, he begins with an example:

Suppose I am engaged in an argument with you about an ultimate moral question; a question about, say, whether we should vote for the libertarian party at some election as opposed to the social democrats…I come to the argument already judging that we should vote for the libertarians, and already motivated to do so as well…let’s suppose you convince me that I am

12 Williams does draw a contrast between moral considerations and considerations of “importance”. The latter consists in questions such as ‘what makes life meaningful’, which must be answered from the first-person point of view, supporting the view that moral considerations are impersonal ones. 1985: Ethics and the Limits of Philosophy. London: Fontana, 1985, see especially pp. 184-186 13 Conversely motive-internalism is the view that there is an internal, necessary connection between judging that X is right or good and being motivated by X. My arguments in this chapter seem to also bear on Smith’s claim that externalism is wrong.

123

ultimately wrong. I should vote for the social democrats, and not just because the social democrats will better promote the values that I thought would be promoted by the libertarians, but rather because the values I thought should and would be promoted by libertarians are themselves ultimately mistaken…what happens to my motivations?14

Assuming fictional Smith (S) is a good person, we need to say how S acquires

a new motivation following this new particular judgment—that social democratic

principles are the right values. Unless we can say that his new particular judgment

itself causes him to be motivated, we will not be able to say that S is directly

motivated by his particular values.15 In order to maintain that there is any connection

between his newly formed particular value and a motivation for this value, we would

have to say that S has a more ultimate underlying concern which provides his

motivation.16 Smith maintains that this underlying concern would have to be a self-

conscious motivation to do the right thing in the abstract.

One can easily object to Smith’s description of the content of the underlying

concern the good person would be said to have. For instance, it is not at all clear why

this concern would have to be a self-conscious one, rather than a deeply held, but

14 Smith, p. 71 15 Again, for the externalist, judgment is separate from motivation. So Smith’s motivation to vote for the libertarian party, on the externalist account, is not a necessary consequence or part of judging that libertarian principles are right. Therefore, when he comes to have a new judgment, now the judgment that social democratic values are right, he may still have the motivation for libertarian values, since judgment and motivation are separate. In such a case we would say that he is motivated to do something he believes is wrong, which is not an accurate characterization of the good person. Thus, we have the problem of explaining how it is that Smith, who is the good person, comes to have a new motivation for social democratic values. The externalist cannot solve the problem merely by saying that Smith’s motivation to vote for the libertarian party is changed to a motivation to vote for the social democrats following his new judgment. For if Smith’s motivation to vote for the libertarian party is due to a direct concern to vote for the libertarian party, then this motivation cannot account for the new motivation to vote for the social democrats. 16 Even though Smith discusses judgments of rightness or goodness and there may be important differences between judging that x is right or good and valuing x, the differences will not be important for what follows. What matters in this chapter is the structure of concern.

124

partly conscious or even unconscious motivation. More substantively, it is also not

clear why the good person must have a concern to do the right thing. There are other

candidates, it seems, for the content of the good person’s concern. Perhaps he is

concerned with being good. Doing what is right might be a large part of what is

involved with being a good person, but it is not ultimately what the good person is

concerned with. There are two differences between this possible description of the

motivational content of the good person and Smith’s characterization. The first

difference amounts to the distinction between someone who is concerned with taking

actions and someone who is concerned with being a certain type of person. A

concern to be a certain type of person does not always translate into discrete actions

and might largely consist of having appropriate responses, emotions and character

traits. A second difference involves the distinction between being motivated by

rightness and being motivated by goodness. While being right does not seem to

admit of vagueness or degrees, being good does. Being good seems to allow for the

possibility of striving or attempting to be or do more, of being better or worse, but

without necessarily reaching a determinate end. Someone ultimately motivated to be

good might simply be someone who is driven to be better, even if only a little bit

better—the psychology of such a person is such that it seems to involve an inherent

understanding of one’s limitations and it suggests someone who is not relentlessly

driven by moral notions. On the other hand, the picture of someone who is ultimately

motivated to be or do right is of one who has a drive for a certain kind of moral

exactitude or meticulousness.

125

Since we can simply take issue with the exact content that Smith has

identified, it is more plausible to interpret Smith’s objection to be targeted not at the

content but at the structure of concern—the fact that the good person’s concern for

his particular ends would be ‘indirect’ and ‘mediated’ by other, more ultimate and

general values. Thus the important implication of Smith’s analysis is that S’s concern

for socially democratic values is indirect, since it is provided by his more ultimate

concern.

But Smith takes this to be a problem. He argues that the good person cannot

be said to be concerned indirectly for the particular things he judges right to do. Just

as good lovers must have a direct, unmediated concern for their loved ones, good

people, he argues, must have a direct, unmediated concern for equality, justice,

honesty, the welfare of people, and other particular things that they believe are right.

He writes that “a view that holds that the good person cares only indirectly for these

things and cares directly only for doing the right thing, is absurd: it elevates a moral

fetish into the one and only moral virtue”.17 By this last comment, it might seem as if

Smith’s argument merely reiterates Wolf’s argument against moral perfection—that a

morally good person would end up being one whose personality is overwhelmingly

characterized by moral concerns and as such would lack the complexity and richness

of a normal, healthy personality. But, importantly, the good person in Smith’s

discussion of externalism need not always be acting for the sake of morality—he need

not experience the psychological and effort demands caused by adhering to simple

consequentialism. Rather Smith’s worry about externalism can be stated as: insofar as

17 Smith., p. 76

126

or when one considers the particular ends that are morally valuable, his concern for

these particular ends is indirect. The only direct concern that the good person would

have under externalism is doing what is right in the abstract. The problem with this

description of the good person for Smith is about the indirectness of his particular

concerns and not about the extent of them or the fact that all of his concerns,

including appropriately non-moral ones, would be subsumed under some general

moral value.18

What this would seem to mean is that even two-level consequentialist views

do not include the proper description of the good person. For even if we take the

ultimate concern to be promoting good consequences, rather than doing the right

thing, and we maintain as Railton does that we can frequently consciously act directly

for the sake of our particular ends, the seeming problem for two-level views is that

they do not contain the appropriate structure of concern, as I will show in greater

detail in the next chapter. Even sophisticated consequentialists must mediate their

concern for their particular ends by appealing to more general values, and this is

especially evident in cases of conflict. However, I will argue next that this indirect

concern is not obviously a problem, as Smith and other authors would have us think.

On the contrary, the structure of concern contained in two-level views is, on

reflection, precisely the sort we would want the morally good person to have.

In the sections below, I do not reject two-level views but instead attempt to

provide a defense of them in response to the charge about the morally good person.

Since Railton tried to defend sophisticated consequentialism by arguing that an

18 Wolf’s central claim can be summarized as following: her arguments “call into question the assumption that it is always better to be morally better.” P. 438

127

adherent can value particular ends for their own sake, the implication of his argument

is that it is despite the two-level structure of concern contained in such views that

sophisticated consequentialists can be committed to their ends. I hope to show that it

is because of the structure of concern contained in two-level views that

consequentialists can be committed to their particular ends.

3.3 Mediating Ends in Our Everyday Lives

Before I proceed, I want to emphasize that problems of incommensurability

between values loom large, and in what follows I do not assume any particular kind

of commensurability between our ends, and especially, any objective

commensurability between our ends as is sometimes assumed in rational choice.19

My analysis instead proceeds from the observation that we do engage in, and must

engage in, comparisons between our ends in many situations. Given this fact, my

concern is how we should consider the comparison of ends and in particular, the

comparison of ends by the morally good person. I will not argue for a specific

standard of comparison or attempt to demonstrate how any standard can resolve the

problem of comparing our ends. Rather, I hope to show here that many of us already

implicitly rely on something like two-level views in trying to compare ends, even if

no easy answer can be found, and it is this trying that is important. Importantly,

though, when we do try to address conflict between ends we both appeal to more

19 Even if we cannot address all of the problems of commensurability, there are a few considerations on this topic that support the arguments here. First, at least some of the work on this topic suggests that even if our abstract values are not commensurable, our particular ends are at least comparable. Furthermore, even in the absence of true comparative evaluative judgment, this would not seem to defeat the claim that we can still try to compare our ends, and it is this claim that is central to the analysis in this chapter. See Elijah Milgram (1997) and Henry Richardson (1994)

128

general values, and treat these values as being more ultimate in certain ways than the

particular ends we are trying to adjudicate.

A discussion by Dworkin demonstrates the problem of comparing our ends.

After claiming that we appreciate the value of art for its own sake, Dworkin asks,

with respect to the value of human life, that “if it is a horrible desecration to destroy a

painting, for example, even though a painting is not a person, why should it not be a

much greater desecration to destroy something whose intrinsic value may be vastly

greater?”20 Dworkin makes two claims here: (a) that we treat both great art and the

life of persons as having intrinsic value in the ultimate value sense; and, (b) that we

can compare their values. Dworkin suggests that art, and presumably persons, also

have value because of their contribution to something else such as pleasure. But even

though we could compare the values of art and persons in terms of their contribution

to something else, it is not this sort of comparison that Dworkin is referring to when

he suggests that the value of persons is vastly greater than the value of a painting. It

is their relative intrinsic values that we are being asked to compare.

Many of us do have the attitude expressed in (a). A problem seems to arise

though precisely because many of us have the attitude expressed in (b). How are we

able to hold both (a) and (b)? For if we are committed to particular ends as valuable

in themselves and not in terms of other values, by what means are we comparing

them? Given that we do make such comparisons, does this mean that we are just

20 Dworkin, Ronald, “What is Sacred?” in Bioethics, ed., by John Harris. (New York: Oxford University Press, 2001) p. 158. He demonstrates the ubiquity of the attribution of intrinsic value by discussing our attitude towards great paintings. We believe that art has its own value apart from its contribution ultimately towards something else, be it our education, pleasure or wellbeing, or anything else. I would add that we furthermore seem to believe that any particular great artwork has value in itself, apart from its contribution ultimately towards the total category of great artwork.

129

choosing at random, merely tossing a coin between which of our ends matter more?

Or do we have some recourse for guiding us through these decisions?

In difficult scenarios, we often appeal to indirect and general concerns to

adjudicate between competing particular commitments, even very personal ones, and

these occasions are much less rare than some authors seem to assume. If we take our

work, our health, our families, and our favorite hobbies to be serious commitments,

and it seems many of us do, then we are forced to choose between these ends very

often. For our particular ends come into conflict frequently, even if we have more or

less settled on how to resolve them. The fact that we do not consciously experience

tension among our different projects and relationships or that the tension does not feel

like an intrusion on our lives does not suggest that our projects and relationships do

not conflict. Often we fail to realize that we are invoking common general values,

but this is because we have adjudicated these sorts of conflicts many times before,

and so solving them at any given time can be habitual. If I am at the library

researching books for a new paper that I would enjoy writing and start to feel that I

am coming down with a cold, I might quickly choose to stop working and go home to

get rest. Even though there is no feeling of real conflict here between my health and

love of my work, that does not mean of course that conflict is absent. We have all

faced a number of decisions such as these in the past and have settled on rules of

thumb for negotiating them—such as, “if I have the luxury of time to take care of

myself when I am feeling bad, I should stop working and do so”. That there is

genuine conflict between my health and my work is obvious in the fact that I would

not always choose my health over my work. If I am at the library finishing a

130

conference paper that is due the next day when I begin to come down with a cold, I

would not choose to stop working. The general value at stake in both cases is simply

that I want a balance in my life between pursuing my work and my health. If I need to

compromise one for the sake of the other, then it should be the one that is least

pressing at that point. Many of us try to balance our disparate ends and internalize this

balance in such a way that we are not constantly faced with the fact that we are

required to reconcile our different engagements. And the balance is achieved not by

appealing to one of our particular ends, but to more general values. Having to make a

decision between two commitments when they conflict on a given occasion can either

reveal more vividly what these general values are or become cause for determining

them.

To see these points, consider the following sort of scenario. Bob has a deep

and rewarding relationship with his wife and loves her dearly. His best friend, whom

he has known for years, is also very important to him. Each of them, on a long

weekend, wants to spend time with him. His wife has just been rewarded at work and

would like to celebrate by taking a small trip, but his friend, who lives in the next

state, has been depressed and calls Bob to request some company for the weekend. In

deciding whom to spend the next several days with, Bob might try to appeal to his

value of attending to the person whose need is greater, and decide to visit his friend.

Not only would this relatively more general concern—attending to the person whose

need is greater—mediate Bob’s decision not to spend time with his wife, but his

action for the sake of his friend would also be mediated by appeal to this concern.

Furthermore, when there is tension between Bob’s concern to attend to the person

131

whose need is greater and some other general concern, such as giving priority to his

wife because of the deep love and respect he has for her, it seems he will need to try

to mediate by a still more general concern than either of these. Bob might feel that

always giving priority to his wife would not allow him to have other rewarding

relationships and this is important not only for a full and varied life but also for the

sake of a healthy marriage, which in turn enhances his life and enables him to better

respect his beloved wife. Bob would be trying to mediate his particular ends in this

case by attempting to find a general value to enable him to decide which particular

end to give priority in which situation.

Again, this is not to suggest that more general values can satisfactorily help us

decide between competing particular ends, but it does seem general values can do so

better than particular ones can. In the example that I just gave, it is not clear how

either of the particular ends—either Bob’s wife or his best friend—is helping him

decide what to do. Nor does it seem likely that some third particular end could help.

How could, for instance, spending time with his daughter, or on his hobby, help him

decide to visit his friend? A relatively more general value can help decide precisely

because it is less contentful—because it relates to both particular ends. The concern

to focus on the loved one whose need is greater relates to Bob’s commitments to both

his wife and best friend because they are both loved ones. What is left for Bob to

determine is whose need is greater, but he will only do this once he has already

invoked a general concern that relates to both ends. This last point about general

values relating to particular ends gives rise to the more important reason, in terms of

the goal of this chapter, for appealing to more ultimate general values in deciding

132

conflict. This reason is that it enables us to take all of our ends seriously and is the

subject of the next section.

It could be argued that I have not demonstrated the importance of two levels

of concern, but only that we need two different kinds or categories of concern. My

examples may only have demonstrated the need to have both general and particular

concerns, with the former having to frequently adjudicate the latter. This kind of

position would not necessarily present a problem for my arguments, since the claim is

that there must be some stable division between general values and particular ends, as

is the case with two-level consequentialist theories, and it does not matter for the

central concern here precisely how we characterize this division. However, in all

such cases where there is a division between general values and particular ends, it

will turn out, it seems, that general values must be in some sense or at a given time

more ultimate than the particular ends if they are to settle conflict in our ends. To see

this, consider the contrary view that would insist that we can bring other general

considerations to help adjudicate conflict in our particular ends without requiring that

these general considerations constitute more ultimate values. But then we must ask

what the status of these other considerations is if they are not more ultimate values.

For it seems that general concerns can only prioritize and adjudicate conflict in our

particular ends if they have a certain priority themselves; a priority over the particular

ones—by being at least in some regards more ultimate values. If the more general

concerns did not have priority in this way, it would be unclear how or why we would

observe or abide by them or why we would let them dictate what activities and

relationships to give attention to, especially given the already assumed importance of

133

these particular ends. This does not suggest that general concerns must always be

more ultimate than particular ends or that there must be a fixed order between general

values and particular ends. Perhaps it is possible to maintain that a general concern is

only held to be more ultimate than particular ones at the moment one is trying to

decide conflict or reduce tension between particular ones. However even if this were

the case, the charge against mediating our particular ends through more ultimate

general values would still apply, if only for that given moment or time that one is

relying on a general value to try to mediate particular ends. I will return to this topic

of the status and ordering of general values in the final section.

I said that as long as there is a stable division between general values and

particular ends, with general ones taking priority over particular ones in certain ways,

it does not matter whether we characterize the account as containing levels or in some

other way. However, there is one kind of account that would seem to be contrary to

my arguments, and it is the view that certain particular ends are constitutive of more

general values or that particular ends are instantiations of more general ones. Joseph

Raz, for instance, seems to hold this type of view when he argues that the claim that

something is capable of being good or valuable for something else is not equivalent to

the claim that it is an instrumental value.21 He writes, “when one says that reading

Proust enriches one’s life one is not pointing to the consequences of the reading.

Rather reading Proust with understanding is such enrichment.”22 The problem with

this view is that it does not seem to help us address the problem of having to trade-off

competing particular commitments. If Proust is enrichment and so is Joyce, then

21 Raz, Joseph. Value, Respect, and Attachment. (Cambridge: Cambridge University Press, 2001) 22 Ibid, p.148.

134

when I want to read both Proust and Joyce but have only time for one, how am I able

to try to use enrichment as a guide for distinguishing between them? It seems

necessary to hold that general concerns are separate from particular ends, that they

are independent enough in order to serve as guides. In that case, when I am pressed

for time, I might decide that Proust would be more enriching than reading Joyce or

that since both are enriching in equally but different ways I will have to read one now

and one later. I might simply toss a coin between them to decide which I delve into

first, but even if I do this, it seems I have come to this conclusion of tossing a coin by

reflecting on, even if only momentarily, a concern to engage in enriching activities,

apart from my interest in either Proust or Joyce. I appeal to the general concern

because I need a value by which to compare the particular ends, but it seems that the

only way I can do this is if the general concern has a certain independence from the

particular ends.

As with the previous example of Bob having to decide whether to spend the

weekend with his wife or with his friend, general concerns mediate the more

particular ones by providing standards by which to compare them, by telling us which

particular end should receive our attention in a given situation and by revealing the

way in which they should be ordered or prioritized. Again, though, it seems the more

general concerns can only do all of this by having a certain priority themselves. But in

order for a general concern to have priority, it must be held as being at least

somewhat independent from the particular ones. The general concern would not be in

a position to operate as an arbitrator of particular ends if particular ones were

instantiations of it or if the particular ends constituted the general one. Even though a

135

general concern can be said to relate to particular ones and to reveal the shared

relationships between them, it must, at the same time, be separate from particular

ends.23

In the beginning of this section, I presented Dworkin’s claims that we both

believe our particular ends are valuable in the ultimate sense and that we can compare

these values, and I suggested that these claims might be inconsistent. While there may

be a way to reconcile them, I hope to show that we do not necessarily have to.

Specifically, we can abandon the belief that particular ends are valuable in an ultimate

sense while still being fully committed to them. As mentioned in the Introduction of

this chapter, even though they are separate issues, a single response might serve in

addressing both (1) the problem of locating the proper source of value of particular

ends and (2) describing how the good person treats his particular ends.24 It seems we

would want to describe the good person as treating his ends to be valuable in the

sense that they are in fact valuable. This suggests that we can determine how the good

person ought to treat his ends only by first determining the source of their value.

Thus, if we hold, as so many authors do, that our particular ends are valuable in an

ultimate sense, then the good person does not try to mediate them through more

underlying general values, for there are no such values available. However I argued

that many of us do try to mediate our ends through underlying general values. If we

do see (1) and (2) as related, which seems appropriate, then an implication of my

argument is to question whether it is the case that the good person’s particular ends 23 One comment related to this point about separateness is that Raz’s kind of view strikes me as insufficiently distinguishing our particular commitments, and therefore as not emphasizing the significance and uniqueness of our particular ends from each other and from general ones enough. 24 I discuss the issue of (1) in greater depth in the next chapter. In addition, I have not of course addressed which values should serve as ultimate ends. This is beyond the scope of this chapter.

136

are ultimate values. But is this necessarily cause for concern? Perhaps the most

important reason for doing (1), attributing the proper source of our values, especially

in the context of the worries expressed over abstract ethical theories, is to ensure that

we take our particular ends seriously. The way in which we can say the two

questions are related then would be to say that the good person must be someone who

takes his particular ends seriously. What we need to determine, then, is what it means

to do this.

In the next two sections, I will argue that two-level views do enable and

indeed encourage us to treat our particular ends seriously—and this is by ensuring

that we consider each of our particular ends and that we do so in a stable fashion. I

will say much more on the topic of ultimate or intrinsic values in the next chapter.

3.4 Considering Each of Our Ends

As discussed above, I am not defending the claim that appealing to general

values satisfactorily helps one solve dilemmas or conflicts in one’s particular ends,

although as I argued before, I do claim that they do so better than particular ends

could. My goal is to instead defend a broad range of views against the charge that in

virtue of appealing to more ultimate general values for the sake of resolving conflict

in one’s particular ends, whether one can successfully do so or not, one does not have

the appropriate commitment to his particular moral ends. In fact, my claim is that it is

by appealing to more ultimate general values, and not in spite of this as some

consequentialists have suggested, that one’s commitments to his particular ends are

deepened and gain stability.

137

To illustrate these points, I return to examples from personal contexts, in this

section, and then examine more obviously moral contexts, in the following. Although

I do not assume any division between personal ends and morality, I hope to show that

even in the very personal contexts, two-level structures to decision-making encourage

a deep commitment to one’s particular ends. There are two reasons for beginning with

personal scenarios. First, many authors, such as Williams, are concerned with the

nature of one’s commitment to personal ends in consequentialist theories and my

analysis will suggest that charges such as alienation from our personal ends are, at the

very least, too strong. Second, and more important for the direct goal of this chapter,

if it can be shown that two-level decision making can encourage a robust commitment

to one’s personal ends, then this point will be even clearer in the sorts of impartial

scenarios Smith has raised in assessing the characterization of the morally good

person—and the success of my arguments mainly depends on the latter.

Recall Bob who has to decide whether to spend the weekend with his best

friend, who is depressed, or go away with his wife who would like to celebrate her

recent success at work. Even if Bob is ultimately unsuccessful in resolving the tension

between spending time with his friend and his wife, the very act of trying to compare

them by searching for a common standard of value encourages him to: (a) appreciate

his ends taken together; and more importantly, (b) consider each end. Taking the first

point, the concern to focus on the loved one whose need is greater or the concern to

have full and rich relationships with others reaffirms the fact that both Bob’s wife and

his best are loved ones and that they each, in their own way, enhance his life, and thus

also brings these separate particular ends into a relationship with one another.

138

Relating each particular end to each other through a general and more ultimate value

seems to embed each end into a framework of things that are important to Bob, rather

than each one standing alone. That Bob’s relationship with his friend is valuable

because it is related to his general value of having a fulfilling life, with many loving

relationships, connects his appreciation for his friend with his other significant

relationships, including with his wife. It connects his commitment to spend time with

his friend to the other things he cherishes. Rather than this seeming alienating or that

he is not fully committed to his particular ends, relating these ends to a general value

and so also to each other connects them together and may bring out their combined

importance. When we have many particular ends that are very significant to us, and of

course many of us do, the appeal to general values can be seen as the recognition that

each and every one of our particular ends is special, unique, and important and that

we have many more than one such end. Mediating particular ends through more

ultimate general values enables us to recognize the totality of our ends, and in this

way, our ends are being treated seriously.

On the other hand, trying to resolve a conflict between ends without any more

general values to serve as common standards would seem to only give rise to

meditating on each particular end in isolation. By treating his commitment to his wife

and to his friend as separate and ultimate, Bob may fail to see how they are related

and therefore fail to see other ends or fail to fully consider other ends that are not

being attended to whenever he decides to favor any given one. This point brings me

to (b), the second and more important reason for the importance of trying to mediate

139

conflict between ends through more ultimate general values—and that is to consider

and appreciate each end.

When critics charge that the appeal to general values causes alienation from a

particular end, they seem to be implicitly making this claim from the point of view of

the end at hand, or the end one is choosing to attend to. For instance, Bob’s decision

to visit his friend by appealing to his deeper value of attending to the person in greater

need, and the further value of having a fulfilling life might leave the critic cold. It

might seem that Bob’s need to justify spending time with his friend by some other

value—other than his friend—diminishes the significance of that relationship, and

indeed this has been the charge. However, what such a critic is failing to focus on is

the forsaken end; the end one has decided not to attend to at that time. Bob’s

forsaken end in the example is his relationship with his wife.

If Bob were to choose to visit his friend simply because it is his friend, then

when his wife asks Bob why he has chosen to spend time with his friend instead of

with her, his response would be that he chose this way because it is his friend.

Consider though how this would make his wife feel. Could she not reasonably assert

that she is his wife and that Bob should feel that the fact she is his wife is just as good

of a reason, if not more, to go away with her for the weekend? If Bob were to

directly appeal to his commitment to his friend then this would seem to leave his wife

feeling that she is not as important to him as is his friend. Furthermore, she would

not be unjustified in thinking that Bob does not seem like he has given her much

thought in this decision. On the other hand if, as in the example, Bob tries to mediate

both of his commitments through his deeper values and concerns, then his response to

140

his wife will be very different. He would say that he has ultimately decided to spend

the weekend with his friend because his friend seems to need him more at the time

than she does. Furthermore, he would say that even though she is his wife, he cannot

always prioritize her because it is important to have many deep relationships in his

life, which will in turn enhance the strength of their marriage and his commitment to

her.

I am not suggesting that the second response would placate Bob’ wife or

would leave her happy to spend the weekend alone. I am instead making the

comparative point that this sort of appeal is more likely to make his wife feel that Bob

values her, because it reflects the fact that Bob considered deeper values in making

his decision not to spend the weekend with his wife—that he did not only consider his

friend. Thus from the perspective of the forsaken end, appealing to more ultimate

general values in situations of conflict implies that the decision maker has, at some

point at least, reflected on and deliberated about both ends, including the forsaken

one, which in itself demonstrates his commitment to that end despite the fact that the

end was not favored in this scenario. Appealing to more ultimate general values, in

the way that Bob has, forces us to stop and examine the loss involved in not choosing

other ends when we decide to favor certain ones, and the importance of this is

difficult to overstate when we have many precious ends.

While a commitment to such a general concern as wanting enrichment might

be more ultimate than particular ends in some regards, this would not imply that one

is not committed to the more particular ends. Rather, appealing to a more general

ultimate concern can be seen as a means for trying to contemplate and balance our

141

many disparate particular commitments. I am reminded of a good scene from a

mediocre movie, Erin Brokovich. In this scene, a character who is a lawyer tells the

character Brokovich that she is taking the legal case on which she is working too

personally. Brokovich emotionally declares, “This is my work! My sweat! My time

away from my kids! If that’s not personal, I don’t know what is!” Brokovich

understands that she has made trade-offs among her particular ends in order to live

more fully. She works long hours and very hard in her job not only to earn more

money to care for her kids, but also to feel that her ideas and skills are important and

respected. By valuing a full life, she has traded-off some of her time with her kids for

the sake of another particular end—the quality of her work. This does not imply,

however, that she does not genuinely care about her kids. Indeed, it is mainly

because she cares so much about her kids that she takes her work so personally—she

needs to feel as though her time away from her kids is worth it.

We often feel as if our particular commitments, such as our families and

projects, are ultimate concerns, with no other concerns lying beyond them, and

certainly nothing general or abstract lying beyond them. We might even say that our

particular ends are ultimate in certain areas or domains of our lives. In the domain of

one’s day-to-day projects, one’s research might strongly override all other

undertakings. But even our most worthwhile projects come into conflict with other

domains, such as when doing research conflicts with one’s health or with duties to

one’s family. We must often weigh and trade against one another those things that

are most precious and special to us and do so by appealing to more ultimate and

general concerns.

142

3.5 Mediating our Moral Values

As with our personal ends, many of our impartial ends, duties, and virtues

come into conflict with one another frequently, and there will be a need to adjudicate

between them. To say that they might conflict is not to require that they are somehow

at odds or in opposition with each other. Indeed, Philippa Foot argues that separate

virtues provide constraints and form to one another. The virtue of friendship, for

example, plays its part “in determining the requirements of benevolence, e.g., by

making it consistent with benevolence to give service to friends rather than to

strangers or acquaintances.”25 So we need not think of particular ends and duties as

being necessarily exclusive of one another. Beneficence need not be at odds with the

duty of respecting others. But the actions required to fulfill one duty might leave

another one unfulfilled, and for deciding between them, we need to resort to values

independent of both. For instance, one might judge that it is right to give money to

charities that help the poor in developing nations, but also believe it is right to support

organizations that try to promote equality for women in Muslim nations. Deciding

how much to give to each, or whether to give a donation to only one cause, would

require one to consider other values and, again, it seems the only way these other

values can adjudicate is if they have a certain priority themselves. As I argued earlier

in the chapter, maintaining that particular ends either constitute or instantiate more

general values would not seem to adequately address the problem of priority among

values.

25 Foot, pp.235-36

143

If we insist, as Smith does, that the good person does not mediate his

particular moral ends through other values, even from time to time, we will have

trouble explaining how he attempts to deal with conflicting ends; how he decides in

the face of conflicting particular concerns. We could of course say that the good

person would then simply be directly motivated by both ends. But we will not be

able to explain how one ends wins out, so to speak. On the other hand, if we describe

the good person as caring about his particular ends but also as being committed to

trying to mediate them via more general and ultimate ones, then we will be able to

say that the good person resorts to these more ultimate concerns when he is faced

with conflicting particular ends, and thus be able to explain how one or the other end,

or a different one altogether, prevails.

A likely response to what I have just said is that the good person need not be

so rational and practical in adjudicating his particular commitments. Even if it is

practically important to hold that when there is conflict between particular ends one’s

concern for them must be mediated by more general values, this does not seem to

square with our intuitions about our typical, everyday moral concerns. I am greatly

upset by harm to animals and it does not feel as if my actions on behalf of animals are

motivated indirectly, mediated ultimately by some general concern. But when I

consider the full picture of what would be involved with being directly concerned for

the welfare of animals along with having a direct concern for each of my other

particular moral ends, it seems more plausible that I relate these particular ends to a

more general value, such as wanting to alleviate cruelty in the world, which may itself

be related to an even more general value. Even aside from immediate conflict among

144

particular ends, it does not seem plausible that the good person would not have more

general values as underlying ones because it is not plausible to think of him as having

particular concerns that are not related to each other in some systematic and stable

way.

To see this, consider again the fictional Smith (S). After discussing the two

political positions with someone else, S no longer judges that voting for the

libertarian party is the right thing to do. He now judges that social democratic values

are right. But what has happened to his concern for libertarian values? If S is a good

person, then by his own criteria, what he cares about when he changes his particular

judgment is precisely what he judges it right to do at that time. But this implies that a

moral concern can completely fall away and another one can take its place as a

particular judgment changes. There are several reasons why this is important. First,

it would not even make sense to say that his new particular concern, now in social

democratic values, takes the place of his old concern, since there is no proper ‘place’

of which to speak. S’s new direct concern for social democratic values is not related

to his, now absent, concern for libertarian values. For under the view that we are

motivated directly by particular values without any more general concerns mediating

them, it seems S’s concerns for the two different values are unrelated. But then, it

seems to me, the question we must ask is how the normative debate in which S is

challenged by his friend influences him to alter his judgment about which party is

right to vote for. If his concern for libertarian values is direct, then there would seem

to be no special reason why normative discussion about the virtues of social

democratic values would be of significance to him and cause him to reject his

145

libertarian values any more than, say, a discussion about the virtues of wine—since

the virtues of wine would be no more or no less distinct from the virtues of libertarian

values than would be the virtues of social democratic values. Merely stating that the

two political positions are inconsistent whereas the virtues of wine and the virtues of

libertarian values are not inconsistent does not resolve the problem. We would want

to say that S’s concern for libertarian values and his concern for social democratic

values are related, that they stem from the same ultimate concern, in order to explain

how his new judgment and motivation come to supplant the old ones. We can

furthermore ask why S should care about whether his libertarian values are right in

the first place unless he has a concern to discover what political values are worth

having.

A second reason why it is important to note that S’s concern for libertarian

values simply ends as he changes his particular judgment is that his values are only as

stable as his particular judgments are. To see why this is a problem, consider the

contrasting view that is being defended here. With a two-level structure of concern,

general concerns would underlie more particular ones. So when S would acquire the

judgment that the social democrats are right, his new concern for social democratic

values would be provided by this new judgment together with, and mainly from, his

already existing, underlying values. A plausible way to think about this is to hold that

his change in particular judgment would refine and provide specific shape to his

existing, underlying general values. On the other hand, under Smith’s view, when a

particular judgment changes the good person would lose a concern altogether, instead

of just the resulting concern that is given its specific character by the judgment. But it

146

seems that this would be inconsistent with how we would want to describe the good

person. We think of the good person as someone who reliably has certain deep and

general concerns and values that transcend particular judgments, not merely a

collection of direct concerns that are only as stable as are his particular judgments

about what is right or good, or kind and compassionate. We do not think of him as

having a series of unconnected direct motivations for particular things like honesty,

equality, and the well-being of his children, without any broad, more coherent, and

more stable way of being concerned for all these things together. More specifically, it

seems that the good person is someone who has stable patterns of particular moral

commitments and having these implies having a way in which a particular

commitment is connected with and fits and works together with other particular

commitments. Having more general ultimate values is important to achieve this, as

the examples from above attempted to demonstrate. The good person need not be so

rational as to have all the details worked out—indeed it seems we would want to

encourage some flexibility in thinking—but it is important that he is willing to

consider how particular ends are or might be related to one another.

This last comment brings me to my final reason why we do not think of the

good person as one who has distinct and direct particular concerns. As I suggested

above, in our personal lives we tend to balance our separate particular ends in a way

in which we are not constantly faced with having to make decisions. Many of us

reflect on and balance our disparate ends and internalize this balance in such a way

that we are not constantly faced with the fact that we are required to reconcile our

different engagements. But more important than someone who cares deeply about all

147

of his personal ends, the good person is someone who periodically reflects on and

compares his particular moral ends, even apart from being faced with immediate

conflict among ends. It might be very important to someone who cares deeply about

his specific personal commitments to reflect on them and try to harmonize them as

much as possible; it would probably enable him to better determine how to go about

his life and help him achieve a greater sense of fulfillment. But one could argue that it

is not constitutive of someone who cares deeply about his personal commitments to

be willing to reflect on all of them. Your commitment to your spouse might trump all

other particular personal commitments no matter what, and you might be unwilling to

hold this commitment up to scrutiny. Similar to William’s example, this might be the

best way to think about your spouse—your unwillingness to question your

commitment in fact may indicate just how deeply you care about this particular value.

Even though this view about our personal ends is debatable—consider someone’s

unwillingness to scrutinize his marriage for fear of discovering that it is not a good

one—this is not how we think of a morally good person. It is constitutive of the good

person to be willing to reflect on and compare his particular moral ends from time to

time, to assign importance to them and consider how much each matters, even if only

to ponder how hopelessly difficult it is to prioritize his particular ends. But if he is

doing this, then at least implicitly, he is trying to rely on more general ultimate

values.

148

3.6 Conclusion

It seems clear that simple consequentialist theories do not reflect the good

person’s dispositions towards his particular ends—it is important for reasons made

clear by Railton, Williams, Wolf and Smith that the good person is not always

focused on general values and, especially, is not always thinking of general moral

values. I have also argued, however, that treating our particular ends as direct and

unmediated is not consistent with the description of the good person either. The good

person must be someone who takes his particular ends seriously and two-level views

do enable and indeed encourage this by encouraging us to consider each of our

particular ends. Without trying to mediate our particular concerns through more

general underlying ones, we would be treating the former as separate and ultimate,

and the problem with this approach is that it can cause us to overlook what other

concerns are being forsaken whenever we abide by or attend to any given one.

Before ending this chapter, is important to address a potential

misunderstanding of the arguments presented here. This can be brought to light by

raising another of Williams’ objections to systematic moral theory and to

consequentialism in particular. Williams argues that it is crucial to avoid sparseness

in our ethical resources—to avoid reducing all of our ethical notions to just one, as is

often explicitly done in rational choice theory.26 My arguments in this chapter do not

attempt to deny this. Appealing to a more general common value to try to mediate

tension between particular ends does not mean that we must reduce them to a single

26 Williams, Bernard. 1985: Ethics and the Limits of Philosophy. London: Fontana, “Our major problem now is actually that we have not too many but too few, and we need to cherish as many as we can” (p. 117)

149

value. Furthermore, I am also not suggesting that we will not face real dilemmas

between our particular ends, for there surely will be tragic conflicts.

I have argued that unless we compare and reflect on our particular ends by

trying to mediate them through more general values, we run the risk of neglecting

some of our most precious ends. However, it is not necessary for the purpose of

mediating and comparing our particular ends to relate all of them to a single ultimate

value. For, importantly, it is not clear that we are merely pushing the problem of

having to compare our concerns back to the more ultimate level. As I have argued,

our more ultimate values must be somewhat general if they are to serve as mediating

and resolving tensions between other concerns. But since they are more general, it

seems they will not come into conflict with each other as frequently or as sharply as

our particular commitments do because, among other reasons, they will not enter our

lives or our societal or institutional practices nearly as regularly. We will frequently

be in the position of having to trade-off particular ends, such as protecting the habitats

of certain animals and promoting the needs of laborers in developing nations. On the

other hand, the values of compassion and of equal respect will less frequently

conflict. But there is a more subtle and important reason why two-level views may

not merely push the problem of having to adjudicate our values back, and this is

because it might not be as morally important to frequently determine how general

values conflict. Because our particular ends are so significant and serious to us, it

seems more important to frequently consider each of them and the relations between

them, and to take general values as fixed foundations for mediating and reflecting on

our particular ones. Consider an imperfect analogy from architecture. It is critical to

150

have a strong foundation to a house. But once a foundation is secure, it seems more

important to attend to the roof, windows, and doors—the very stuff of the house, the

parts of the house that are most significant for our everyday living. On any given

occasion, failing to consider the foundation is unlikely to have dire consequences

even when there are problems with it, whereas failing to consider a broken window

could be disastrous. From this last point it could be suggested that precisely because

our particular ends are so important for our lives, we would not want them to form the

foundation of our values.

151

Part I I Rational Choice and Consequential ism

4 Cost-Benefit Analysis in Ethical Decision-making

4.1 Introduction

In the previous chapter I presented the charge that consequentialists who do

not place intrinsic value on their individual projects and relationships are alienated

from meaningful aspects of life. In this chapter I, will discuss how similar sorts of

charges have explicitly been made in the context of rational choice theory—and this

is against the use of cost-benefit analysis in moral decision-making. Cost-benefit

analysis does not always assume that the ends or values to be compared are

consequences, but does possess the mode of ethical thinking I described above. In this

chapter, I will argue that while traditional cost-benefit analysis that relies on some

notion of efficiency, such as pareto efficiency, is subject to a wide range of moral

objections, cost-benefit analysis as conceived as a framework in which to compare

our moral ends encourages broad reflection and the consideration of each alternative

value. David Schmidtz has made a similar argument regarding cost-benefit analysis,

however my position diverges from his in significant ways. Furthermore, I will

argue that far from calling for the maximization of any value, cost-benefit analysis

simply calls for the minimization of costs, or more minimally, the assessment of

costs—and costs are not strictly monetary ones but are rather the values and ends

which are not favored whenever an action is undertaken for the sake of some other

end or value.

152

The context for my discussion on the use of cost-benefit analysis in ethics is

bioethics. Since many bioethical positions and policies must weigh the value of life

or particular lives against other, sometimes more general values, they can be

described as failing to treat life as always intrinsically valuable. So it is especially

fruitful to provide examples and illustrations from bioethics to motivate the

discussion. I argue that far from causing estrangement from our significant values,

the broad sort of cost-benefit analysis advocated here allows us to consider all of our

important values and, in this way, take our values seriously.

4.2 Intrinsic Value and Alienation

The area of bioethics is laden with claims about the intrinsic value of life,

treating persons as ends in themselves, and bestowing priority to individual rights and

autonomy. For instance, in What is Sacred, Ronald Dworkin presents an analysis of

what he believes is the deeply held conviction in the “sanctity of life”.1 Rather than

examining his particular claims regarding how various sorts of life and death

practices should be considered, I wish to examine his characterization of life as sacred

and the implications that follow. Creating a general conception of the sacred, as

presented in the previous chapter, Dworkin argues that “much of what we think about

knowledge, experience, art, and nature, for example, presupposes that in different

ways these are valuable in themselves and not just for their utility or for the pleasure

or satisfaction they bring us. The idea of intrinsic value is commonplace, and it has a

1 Dworkin, Ronald, “What is Sacred?” in Bioethics, ed., by John Harris. (New York: Oxford University Press, 2001)

153

central place in our shared scheme of values and opinions.”2 He tries to demonstrate

the ubiquity of the attribution of intrinsic value by discussing our attitude towards

great paintings. We are said to appreciate the inherent quality of art, and not simply a

painting’s contribution towards our education or pleasure. He then argues:

If it is a horrible desecration to destroy a painting, for example, even though a painting is not a person, why should it not be a much greater desecration to destroy something whose intrinsic value may be vastly greater?3

As presented in the previous chapter, Dworkin makes two claims in this

passage: 1) that we treat both great art and the life of persons as having intrinsic

value; and, 2) that we can compare their values. Again, it seems many of us do have

the attitude expressed in (1) and a problem arises because it seems many of us also

believe that (2) is possible. If we are committed to certain values as intrinsic ones—

that is, valuable in themselves and not in terms of any other values, considerations

and reasons—how are we able to compare them? Given that we do make such

comparisons both in our personal lives and as a society, does this mean that we are

choosing randomly?

To answer these questions, we need to understand precisely what is meant by

the characterization of treating something as intrinsically valuable.4 Unfortunately,

Dworkin fails to offer any specification, and indeed it is difficult to do so since the

very nature of the concept, as something that is good in itself and not in terms of

2 Ibid., p. 158 3 Ibid., p. 161 4 In this paper, the notion of intrinsic value, valuing something as an end itself, is being contrasted with instrumental value, valuing something as a means. That is, what is of concern is the way in which we value things, as opposed to the way in which things have value.

154

anything else, seems to defy specification. However, other philosophers have

attempted to give more content to the concept. Therefore in order to flesh out

Dworkin’s view, it is useful to examine one such account. As discussed in the

previous chapter, Peter Railton offers an account of what it means to treat something

as intrinsically valuable and does so more clearly than most philosophers.5 For

Railton what it means to be committed to an end or value in an intrinsic, as opposed

to an instrumental, manner is a “matter of (among other things) whether it furnishes

one with reasons for acting that are not mediated by other concerns.“6

Railton maintains that being committed to a value in an intrinsic manner

means that it gives us a direct, unmediated reason to act. Furthermore, he states that

this reason need not always override other reasons for acting, for otherwise we would

be able to have only very few such commitments. My interest, though, is in cases of

conflict between two or more such commitments, what is furnishing us with reason to

act. We could say that one’s particular values are still providing unmediated reasons

for acting in these circumstances. However, it would not seem to satisfy Railton’s

characterization of someone who is committed to particular ends to merely maintain

that they provide a reason for acting, even any weak reason, without also adding that

these reasons motivate one to act for the sake of those ends. In order to count as

something that we hold dear, it is not sufficient that a particular value merely

furnishes any unmediated reason for acting. An alternative and plausible way to

interpret Railton’s criterion, especially given his phrasing throughout that we act for

5 Railton, Peter. “Alienation, Consequentialism, and the Demands of Morality,” in Consequentialism and its Critics. Ed by. Samuel Scheffler. (New York: Oxford University Press), 1988 6 Ibid., p. 101

155

the sake of our particular ends, is to say that we are motivated or concerned to act

directly for the sake of our particular ends.7 I am not suggesting that we read

Railton’s claim as being that we must always be motivated directly by a particular

end in order for that end to count as an intrinsically valuable. That is too strong a

position for any plausible view. Rather the claim should be read as maintaining that

when we are acting for the sake of a particular end, our motivation for doing so is an

unmediated one. But an example will demonstrate how a sophisticated

consequentialist’s motivations on behalf of his particular ends will need to ultimately

be mediated through other concerns—those that form the level of good consequences.

Say that a sophisticated consequentialist regards both his best friend and his

family as intrinsic ends in Railton’s sense that he acts directly for the sake of each.

However, when these ends come into conflict, he chooses to act for his family. On

what basis does he choose his family? He might choose his family because, to use

Railton’s terminology, his family is an overriding end. If his family is an overriding

end, however, then at least implicitly, he is comparing his ends. How else can he

judge that his commitment to his family is stronger than his commitment to his

friend? If he is comparing these commitments, then it seems his concern for his

family and his concern for his friend are mediated ultimately by other, more general

ends. Perhaps he loves his family more and because of this feels a greater sense of

duty to them. But if he is a good sophisticated consequentialist, then he must be

choosing to respond to or honor his greater affections to his family, ultimately, by

appeal to the general level of good consequences. For even if this level is partly

156

constituted by the end of devotion to family or loved ones, it would be implausible to

assume that his family and friend constitute good consequences without the theory

losing its consequentialist character. As I discuss below, there are also further

problems with viewing particular ends as constituting more general ones that are

independent of the contents of consequentialism.

A second reason he might choose to act for his family over his friend is

because circumstances simply favor his sacrificing the time that could be spent for his

friend on helping his family. For example, if his friend is on the other side of the

world and his family is living nearby, then it will simply be easier, less costly, to help

his family and he might therefore be able to help them a lot more. However, the issue

of comparability is even more apparent here. When required to make choices between

equally held commitments, the sophisticated consequentialist must adjudicate

between them based on concerns that are independent of both. When there is conflict

between his more particular ends, none of which is overriding, his concern for acting

will even more obviously be provided by none of them. He chooses to act for his

family because he can do more to help them, but this motivation to help more where

he can is not a direct concern for his commitment to his family. He must resort to

considerations indirect to, external to, both his family and his friend to decide

between them. While he may frequently be consciously motivated by his family and

by his friend, he must ultimately, if he is a good consequentialist, mediate these

motivations by appealing to the outcomes produced by his decisions. Examining

cases of conflict makes this point clearer, but the point does not apply only to

157

scenarios in which two or more commitments are in immediate conflict, as I

discussed in the previous chapter.

I now wish to add that perhaps one way to think about our commitments is to

say that they are often weak intrinsic values. They provide reasons for taking some

action, but fail to provide reasons for not taking that action. For instance, say that my

commitment to my friend furnishes me with a reason for comforting her when she is

depressed—call it action X. When X conflicts with Y, the action I would undertake

for reasons from my commitment to my work, after some point of engaging in X, I

choose to do Y. If I choose to no longer do X because of my pressing need for Y,

because I am on the brink of getting fired, my commitment to my friend is failing to

provide the reason for not doing X. Some other commitment supplies my reason for

not doing X. On the other hand, if my reason for no longer comforting my friend,

doing X, is because it seems to be rendering her dependent on me to the point that she

is becoming more vulnerable, then my commitment to my friend does supply my

reason for no longer doing X. We could describe this as a strong intrinsic value, but

only at that time.

It is implausible to think that my commitment to my friend or any of my other

particular values would always be a strong intrinsic value. Each may indeed

constitute a strong value in some local sense or in some domain of my life. For

instance, in the area of my day-to-day projects, my research might strongly override

all other undertakings. But even my most worthwhile projects come into conflict

with other values, such as when my research conflicts with my health or my duties to

my family. We must often weigh and trade against one another even the things we

158

think of as having intrinsic value, and do so by mediating them through other, more

general, values. When we are mediating them through other values then, as I

demonstrated in the previous chapter, it is not true that we are making decisions on

behalf of our particular values for reasons that are internal to those values. To put

this point in terms of the contrast between weak and strong intrinsic values, we can

say that strong intrinsic values are only valuable in themselves, but this would not be

an accurate description of weak ones.

The end of maximizing good consequences can now be viewed as an

intrinsic value in the strong sense, but it does not commit us to disregarding our other

serious values. Rather, appealing to the strong end of maximizing good consequences

is a means for reflecting on and balancing disparate, strongly-held values. More to the

point, serious commitments can provide reasons for acting—if only in the weak sense

as described above. Each of our serious particular commitments cannot plausibly

always be an intrinsic value in the strong sense even for non-consequentialists. Our

commitments, if we have more than one, will sometimes give reasons for acting not

mediated by other concerns, but even a consequentialist can have intrinsic values in

this sense.

When two serious commitments conflict, having to choose between them may

be difficult to say the least, but comparing them by resorting to external reasons does

not render one cold, indifferent, or even alienated. Rather, it is an inescapable burden

of the condition of scarcity. The main difference between most consequentialist

theories and most non-consequentialist theories in terms of the way in which

commitments are viewed is that in the case of consequentialism it is decided prior to

159

serious clashes of commitments what would be a general reason for choosing one

commitment over another.8 I doubt that anyone would want to fix a charge as

serious as alienation on when in the decision-making process a standard for choice is

invoked. In fact, as suggested in the previous chapter, deciding on a general value for

adjudicating particular ones in advance forces one to consider how each particular

value might be related to other values and thus encourages broad reflection one’s

values even in the absence of obvious conflict between values.

To revisit Dworkin’s claims about the intrinsic value of life in light of what I

have argued above, it seems that we in fact treat persons as intrinsic values in the

weak sense, given that we often choose not only between persons and other values

that we feel are very important, but also between persons and other persons, and must

do so by appeal to external reasons. Even Dworkin implicitly provides a reason for

the intrinsic valuation of art and human life, a reason which is itself a value. The

reason, he argues, that we value art and human life is because of the assumed creative

and historical processes that give rise to such objects. Thus even aside from their

contribution towards utility and pleasure, it is not the case that the value of human life

lies only in itself. We can ask why human life is so valuable, and one reason that is

provided is because it is the product of either natural or supernatural creative forces of

evolution.9 Rather than breeding alienation and coldness, an answer such as this

seems to connect us to the world and the individuals in the world.

8 But note that this does not mean that there is only one measure of what constitutes good consequences. 9 Dworkin, pp. 162-164

160

Perhaps one reason we feel very strongly that certain objects, including life,

are valuable only in themselves is because the connection between them and the

external reason for their value is not always obvious and direct. Another reason for

the attribution of strong intrinsic value to a certain object might be because the way in

which it realizes a more general consideration is significant, or because the

consideration that it realizes is a particularly important one. However, it is important

for the sake of making decisions to recognize that if we have more than one thing of

which we speak of in terms of intrinsic value, we will need to look to other

considerations to choose between them when they clash. Contrary to the view that

many economists hold, though, this does not imply that it must always be the same

considerations. Balancing our values does not require that there is a single ranking, or

indeed any fixed ranking at all, of our values. How we judge which values are more

salient than others can vary with the circumstances of a situation as the particular

conditions and circumstances will warrant that certain considerations be more

prominent. Nor is it necessary, again contrary to many economists, to reduce all of

our values to a single value—even many consequentialists recognize this, as they

frequently describe the state of good consequences as consisting of many different

values. But most importantly, weighing and balancing our ends by mediating them

through other general considerations does not imply the diminished significance of

the things we take to be sacred.

In concluding this section and for the purpose of bringing the discussion to

bear explicitly on bioethics, I would like to stress the following points: (1) as a

society we do not in fact view life as sacred, if believing life is sacred entails that it

161

offers us reasons for acting in the strong sense; (2) we simply will not be able to hold

each life as sacred, given the condition of scarcity; and, (3) holding any life to be

sacred, given the condition of scarcity, will hinder us in considering other values and

thus prevent us from being able to fully consider all of the options available for a

policy or action, including those that are not direct alternatives. The belief that

something is intrinsically valuable in the strong sense and is to be protected at all

costs can create a different kind of alienation—distancing us from other important

values. This last point will be the emphasis of the next section.

4.3 Comparing Values in Bioethics

Railton’s argument that consequentialists can be alienated from rich, unique,

and beautiful aspects of life sounds all too familiar to economists and other who

employ cost-benefit analysis. Cost-benefit analysis (CBA) has traditionally been

conceived of as the attempt to ensure that the benefits of any action or policy

outweigh the costs of that action or policy. Many ethicists greet the option of

employing cost-benefit analysis (CBA) with repugnance. How can we compare our

most precious values to costs? How are we able to place a monetary value on the

ephemeral wonders of the environment, on the splendor and richness of art, and

especially on the sacredness of life? How can we price that which is priceless? This

kind of questioning often leads to the conclusion that CBA and those who employ it

are indifferent, and worse, heartless—alienated from what is most valuable in life.

But I will demonstrate that there are different and more important reasons for

162

employing CBA—in its broadest formulation, CBA involves comparing values

against other values and doing so by appealing to still other, more general reasons.

Thus, CBA can be viewed as a form of consequentialism, without necessarily

maintaining that there is only a single strong intrinsic value and, furthermore, that this

value is good consequences such as welfare or happiness. Since broad CBA calls for

comparing and balancing values with other values, it is susceptible to the charge that

it fails to treat our values intrinsically and thus causes alienation. In this section, I

will continue my argument against the claim of alienation in the context of CBA.

To motivate this discussion, it will be useful to first demonstrate a seemingly

common misperception involved with the use of CBA, one that does not extend to

consequentialist frameworks in general. Some ethicists hold the vastly mistaken

belief that when CBA is employed, we are reducing values to “economic values”. In

Which Way Down the Slippery Slope, Ruth Macklin epitomizes this misunderstanding

when she discusses the possibilities of sliding down a morally problematic slope from

making decisions on euthanasia in our society towards the Nazi euthanasia program.10

She argues that contrary to others who believe the most dangerous practice is

withdrawing treatment from patients who lack the capacity to decide, the greatest

danger “lies in the use of an economic rationale”. “When the justification offered for

terminating treatment is that it is not ‘costworthy’, or that it is consuming a

disproportionate amount of society’s or an institution’s resources, the slide down one

of the slopes to the Nazi program has already started”. 11

10 Macklin, Ruth, “Which Way Down the Slippery Slope? Nazi Medical Killing and Euthanasia Today,” in Bioethics, ed., by John Harris. (New York: Oxford University Press, 2001) 11 Ibid., p. 130

163

But discussing the costs of any policy or action, and employing CBA more

generally, is not about monetary value. It is not inherently about anything. It is

simply a means for comparing our values and the costs of achieving those values.

Furthermore, even when CBA is presented in the terms of economic value, it still is

not inherently in terms of anything in particular. Monetary measures are just that—

measures. When we discuss the monetary cost of terminating life-supporting

treatment to a patient for whom doctors will not be able to restore functioning, it is

not necessarily the case that the money will otherwise be used to line the pockets of

hospital CEOs. Discussing the monetary costs of withdrawing treatment can be

shorthand for saying “if we continue to sustain this patient, we will not be able to

spend that money on better patient care for other elderly patients, longer post-delivery

hospitalization for new mothers, or free immunizations for area immigrants” since

there are limited resources. CBA is a means for comparing uses of monetary

resources against other uses of monetary resources, not for comparing important

values against monetary resources.

David Schmidtz makes similar points.12 Against the background of using

CBA for environmental concerns, Schmidtz argues that CBA is a method for taking

our values seriously. For Schmidtz, the strongest argument in favor of employing

CBA is that it is a means for introducing accountability. Part of what it means to do

proper CBA is to account for all of the direct costs involved with any option,

12 Schmidtz, David. 2001. “A Place for Cost-Benefit Analysis,” Philosophical Issues II, 148-71. Blackwell Publishers, Inc.

164

including the costs that are imposed on others, or the external costs.13 Often the

external costs of an option will be so great, that it will not pass the cost-benefit test,

and this implies there is a means for holding people accountable for their decisions

involving important values. Many of the points Schmidtz makes in his discussion of

CBA are very good, and so I reiterate and expand upon them here. But I also argue

that there are important divergences between his view and mine, noting when he fails

to go far enough in his defense of CBA, and conversely when he goes too far.

To begin with, both Schmidtz and I maintain that there are two main reasons

for employing CBA, although I will argue below that my conception of these reasons

allows us to employ CBA to reflect more broadly on our values. These two reasons

are the following:

(1) To examine which means of protecting our values will be most cost-effective

and

(2) To explore what important values might be forsaken when favoring others.

The point of the first reason is that it is of course better if we are choosing

efficient means for achieving whatever we identify to be our goal, purpose or end.

Even if a particular value, such as life, is taken for granted as our end, we still need to

consider different means of protecting it. Some means will be more cost-effective

than others and so will enable us to either protect more life or instead divert the spare

resources to the achievement of other values. This latter possibility raises an

important consideration often misunderstood by philosophers, perhaps even by

13 To give an example relevant to bioethics, when a woman is considering the costs associated with terminating a pregnancy, the costs to the father might be considered an external cost, both in the case that she decides to continue the pregnancy and in the case that she decides to terminate the pregnancy.

165

Schmidtz. This point is that even though cost-benefit analysis employs a maximizing

rationality, this does not imply that our aim must be to maximize some value.

Schmidtz argues that even if CBA considerations call for us to maximize a value, this

would not imply that we ought to do so. Sometimes, he argues, such as when

confronted with the possibility of being able to save five patients by providing them

the organs taken from one patient, we should merely respect a value rather than

promote or maximize it.14 Respecting, as opposed to maximizing, life would imply

that we should not sacrifice one patient to save five. But Schmidtz mistakenly

believes that it would be despite CBA, and because of other moral considerations,

that we should resist maximization. However, CBA, outside of its use in the context

of any particular moral theory, only implies that our aim is to minimize some cost, not

to maximize some value.15

Importantly, exactly what we wish to achieve by minimizing costs is not

entailed by the use of this kind of maximizing rationality. It is entirely up to us and

will depend on the values we take to be most important. By minimizing costs, the

saved resources can be applied either towards securing more of the same value, or of

fulfilling some other value. For example, if our end is to secure the procurement of

organs and tissues, a default policy that treats all people as potential donors upon

death, unless people actively opt-out, might be a considerably lower cost means than

using animal organs, which is often ineffective and raises a number of other moral

considerations involving the status of animals. Also, preventative care may be a far

14 Ibid., p. 482 15 Allen Buchanan has suggested to me that it might be viewed as having an even more minimal function: to simply allow us to know what the costs of any action are.

166

more cost-effective way of prolonging life than life-sustaining treatment. In both

cases, with the resources that are saved we can either spend them on achieving more

of the same end, or divert them towards other values.

The other reason for employing CBA, the second reason stated above, is that

it forces us to stop and examine the costs of not choosing other values and ends when

we decide to favor certain ones. In other words, it is a method of comparing not only

ends with means, but also of comparing ends with other ends. Schmidtz also makes

this point, and in making it, he implicitly invokes the notion of opportunity costs, but

he does not go far enough in his position. He argues that our values will sometimes

come into conflict and we will sometimes be forced to compare them and choose

between them. There are a few things to say in response to Schmidtz’s

characterization of the problem. First, because we live in a world with limited

resources, and indeed simply because time is limited, whenever we choose some

policy other opportunities are sacrificed. If we take opportunity costs seriously,

whenever we decide to apply resources to some end, not only do we need to consider

the obvious alternatives to those resources, those alternatives that are in direct

competition, but we should also consider the less obvious alternatives. We should

resist the quixotic temptation for believing that those problems that are immediately

before our eyes are the only problems that merit our moral attention. For instance in

taking the action of prolonging life sustaining treatment for some patient, the

opportunity costs are not only the costs of not prolonging life sustaining treatment for

other patients. Some of the other forgone opportunities that are not as obvious might

be better patient care for other elderly patients, longer post-delivery hospitalization

167

for new mothers, or free life-saving immunizations for area immigrants. These are all

important considerations and so illustrate the need to take (all) forgone values

seriously when making decisions about actions or policy. Importantly, though, there

is nothing in the concept of opportunity cost that implies that competing alternatives

need to be compared based on questions of efficiency. Another policy might be

considered on grounds of what is more fair or equitable, or any number of other

values.

A second thing to notice about how Schmidtz characterizes the problem is that

he seems to believe that what we are forced to compare are the means or methods

employed to bring about intrinsic values. They are, in other words, relatively

instrumental values that Schmidtz sees as being the objects of comparison and

decision-making. For instance, his approach appropriately calls on us to think about

the costs and benefits for pollution when considering a particular method of

alleviating pollution, such as recycling. But it does not challenge us to compare the

value of alleviating pollution with other values. Thus, Schmidtz focuses on

employing CBA as a method of weighing the direct costs and benefits of a given end

or value, like alleviating pollution, and his view does not encourage broad reflection

and comparison between competing intrinsic values. A potential reasons he places

the emphasis on direct costs, as opposed to the true opportunity costs, of any policy

might be that he views values as lying outside the framework of cost-benefit analysis.

Recall that Schmidtz suggests that respecting life would call for us to refuse

sacrificing one patient to save five. But, as I discussed before, Schmidtz mistakenly

believes that our refusal to sacrifice any lives would be despite CBA, and because of

168

other moral considerations. He views our values as being moral, and therefore,

alternatives to CBA, implying that CBA is not about or at least not always about

moral values. If this is one’s view, however, then an important question to ask is how

we are able to justify the implementation of CBA into any area of ethics. Unless

CBA is itself a (thoroughly) moral framework, what warrants its use in ethical

decision-making? More to the point, if one sees the comparison of our moral ends as

causing alienation from our ends or as causing other significant moral problems, then

the employment of something like CBA would seem to compete with taking our

values seriously, or with respecting life in Schmidtz’s example. But I argued above

and in the previous chapter that there is no good reason to view consequentialist

frameworks in this way.

We are often forced to compare our most precious values. This is an

unfortunate reality of the scarcity of means and time. Since there are limitations on

the resources that we possess to employ towards the realization of our valued ends,

some kind of decision-making approach is often needed. Far from reducing the worth

of our ends, though, CBA is the enterprise of taking our ends seriously, when there is

more than one intrinsic value. It is a means of ensuring that our ends are realized at

the lowest costs and that the alternative values are considered, and both purposes

speak to the fact that there are many ends, and even many quantities of the same end,

that we deem important.

169

4.4 Beyond Alienation-- Other Costs and Benefits of using CBA

A significant advantage of employing CBA as a framework for comparing and

balancing values is that it does not require the use of any particular moral theory. Its

application can be integrated into any substantive approach. On the other hand, other

recent attempts, especially in the field of bioethics, at providing a framework for

comparing our values may not so easily be reconciled with all moral theories. Most

notably, Tom Beauchamp and James Childress have advocated the use of general

guidelines and standards, in the form of principles, to address pressing questions in

biomedical ethics.16 Their view of principles is the following:

Principles are always binding unless they conflict with other obligations. When a conflict of norms occurs, some balance, harmony, or form of equilibrium between two or more norms must be found; or, alternatively, one norm overrides the other.17

When there are conflicts, they propose a model of balancing the principles for

similar reasoning employed in the advocacy of CBA in this paper: “A model of

balancing keeps options open without ‘flatly prohibiting’ them.”18 But because they

employ four principles as the basis of many more specific rules, without the

background of any comprehensive ethical theory, there is a greater likelihood that

their approach, rather than the kind of broad CBA argued for here, will be rejected by

the advocates of particular ethical theories. Namely, because it provides content,

16 Beauchamp, Tom L. and Childress, James F., Principles of Biomedical Ethics, 4th ed. (New York: Oxford University Press, 1994) 17 Beauchamp, Tom. “Principlism and Its Alleged Competitors,” in Bioethics, ed., by John Harris. (New York: Oxford University Press, 2001), p. 481. 18 Ibid., p. 483

170

rather than merely being a framework for considering any different values, their

version of Principalism may more easily come into conflict with theories that disagree

over the specific principles chosen, or that reject the emphasis placed on only these

principles or the reasons for which the principles are important.

Despite the fact that cost-benefit analysis can be construed as the call to

consider all of our values, many philosophers and others find that the traditional use

of CBA is morally problematic in many ways.19 Historically CBA required only a

Pareto improvement, which occurs if someone is made better off without making

someone else worse off. This appears to be a non-controversial, minimal ethical

requirement that escapes the problems of interpersonal comparisons. If our federal

government were to receive a windfall sum of money and could devote those

resources to cancer research without thereby lessening the amount of resources

directed towards AIDS research, this would seem to be very appealing. However,

there are a number of problems with the notion of Pareto improvement, and while

some of them are too involved to discuss here, the most significant problem can be

summed up briefly: Pareto improvements are rare. Economic changes usually

involve winners and losers.

To address this very serious shortcoming to Pareto improvements, economists

developed the idea of a potential Pareto improvement, which is when the winners can

compensate the losers and still be better off than before. In theory, one asks the

“winners” how much they would be willing to pay to have a policy and the “losers”

how much they would require for compensation in order to accept the policy, with 19 For an excellent critique of traditional CBA, see Hausman, Daniel and McPherson, Michael, Economic Analysis and Moral Philosophy. Cambridge: Cambridge University Press, 1996, pp.93-99.

171

how much one is willing to pay depending upon how strong one’s preference is. In

practice, economists use various techniques to infer people’s preferences from their

economic choices. The charge that CBA ignores issues of justice and fairness mainly

enters with the use of potential Pareto Improvements and the corresponding use of

“willingness-to-pay”. To start with, an analysis that relies on “willingness-to-pay”

(or “required compensation to be paid”) might be too centered on the preferences of

rational humans and would likely exclude animals, infants, the cognitively impaired,

and the environment. Second, “willingness-to-pay” often depends on how much one

can pay, and thus how much money one has, and on one’s idea of what would be an

appropriate price to pay. If a particular policy benefits the poor and harms the rich,

what the poor would be willing to pay for the new policy might not be enough to

compensate the rich, in which case the policy would not pass the Pareto test. Even

though it might not have any net-benefit in this sense, we may still be convinced that

the poor would benefit a great deal more than the rich would lose from the policy’s

implementation.

Third, and perhaps most importantly, potential Pareto improvements ignore

questions of fairness: the compensation considered is only hypothetical, and in

reality, some will win and others will lose. If over the long run each person wins as

often as she loses, then that a particular policy benefits some and harms others might

not render it unfair. But there is a bias in CBA against the preferences of the poor

and other beings besides rational humans to whom we would want to confer moral

status. It is also questionable whether policy or actions should be based on inferences

drawn from people’s economic choices because this mechanism would leave out

172

perhaps our most valuable preferences. Some of our most valued “preferences”, if

they can be called so, such as a preference for lives free of suffering, are difficult, if

not impossible, to signal when one buys clothes, gas, and groceries. Furthermore and

perhaps most importantly, even our most valued preferences can change upon the

deeper reflection and criticism that comes from public debate. So CBA should not be

a substitute for public deliberation. Traditional cost-benefit analysis leaves

individuals’ values secure from challenges.

Although I believe that many of the objections raised against CBA are valid, it

is important not to confuse traditional CBA with the kind of broad CBA advocated in

this paper. However, there are certainly limitations to this approach as well that arise

precisely because it does not provide specific content to the costs and benefits

considered, and here is where my position most clearly diverges from Schmidtz’s.

Schmidtz argues that CBA is not a sufficient condition for any decision.

Demonstrating that the benefits outweigh the costs will not render an option

decisively correct because, he argues, other considerations will matter. For Schmidtz,

CBA is a necessary part of decision-making. If an option fails to pass the test of

CBA, or if the costs outweigh the benefits, he asserts it is clear that “further

discussion is not warranted”.20 This at first appears to be a modest position. If the

costs of an option are greater than the benefits, then how can we seriously consider

such an option? However, without giving further content to the “costs” and to the

“benefits”, even this proposal goes too far.

20 Ibid., p. 481

173

Other considerations will matter, of course, and therefore he is right to argue

that CBA cannot be a sufficient condition. But other considerations will also make it

plausible to remove the status of necessary condition from CBA. If the importance or

weight of the costs and benefits are not included in the analysis, then CBA should not

even be a necessary condition. For example, if we are considering expensive life-

extending care for a late-stage terminally ill patient, should we abandon such action if

the costs outweigh the benefits? This is not easy to answer and will depend on

whether the benefits, or values favored, are more or less important than the costs, or

the forsaken values. The important question to ask is: what are the alternative uses of

resources to sustaining life-support for this patient? If the alternative is that many

more lives can be saved through free immunizations, then the cost of sustaining life-

support for the patient is potentially great. In this case, that the costs outweigh the

benefits might imply that further discussion is not warranted. But if the alternative

use of resources is to re-carpet a floor in the hospital, and this would have the result

that a greater number of people would be happier, simply that the costs outweigh the

benefit would not seem to imply that further discussion is not warranted. Even if the

losers are losing more than the winners are gaining this would not be decisive,

because what the losers are losing might not be as important as what the winners are

winning. The full accounting of costs and benefits will depend on what we consider

to be the costs and benefits and how different reasons should weigh in on determining

how important each is.

The difficult task of course is how to consider the importance of various

values. But it is not whether we should engage in cost-benefit analysis at all or ask

174

what the costs are. Above I discussed the flip side of maximizing any value, and that

is minimizing any cost entailed in achieving a valued end. Again, attempting to

minimize a cost of attaining a certain end in the field of bioethics would not imply

that the saved costs would line the pockets of policy-makers or hospital board-

members. By minimizing costs, the saved resources can be applied either towards

securing more of the same value, or of fulfilling some other value. Minimizing the

costs, or more generally, inquiring into the costs of any action or policy is the crux of

cost-benefit analysis that should be embraced in the context of any ethical decision-

making, including in bioethics. We must be willing to meet the actuality of scarcity

candidly and be willing to discuss costs in order to treat all of our values seriously.

Holding life, for instance, as sacred in the sense of always treating it as a strong

intrinsic value, one that cannot be compared, and as something to be protected at all

costs, might prevent us from being able to take other values, and even other lives,

seriously. The unwillingness to balance our values with other values can occasion a

different form of alienation from what Railton has illuminated. By faithfully

adhering to the intrinsic moral status of certain values, we become estranged from

some of the reasons for our dearest values and hence from other important values. We

run the risk of living parochial lives that are detached from others in the world for

which our neglect of comparing and balancing values has significant consequences.

175

4.5 Conclusion

The ideals of the intrinsic value of life, respect for individuals, and treating

people as ends in themselves are important because they signify a society’s

unwillingness to promote the common good at the expense of the individual. They

also place emphasis on values, in particular respect for autonomy, that have

historically been undervalued by various health-care practices. Furthermore, they

offer goals to aspire to. On the other hand, these values must also be vulnerable to

evaluation, criticism and comparison with other significant values. Although this

may seem obvious to some, few authors in the ethics literature seem to fully

appreciate the implications of the condition of scarcity. This is something which

economists have taken seriously. Given that resources and time are both limited, we

are often, indeed, always according to some views, forced to choose between

competing values.

However, related to the problem that CBA might be used as a test for any

decision without full consideration of the significance of the costs and benefits, it

seems that philosophers and others are right in at least one aspect of their objection to

the use of CBA, even of the broad kind advocated here: that those who employ it

seem too easily to want to employ it, and something may be lost in the facility of its

employment. In particular, it seems there is some real import in using the language of

intrinsic value, pricelessness, and incomparability for those things that have

especially high value to us, such as life. Not only might it thwart becoming hardened

to the fact of scarcity and its corresponding implication that we must, unfortunately,

176

make choices between our highest values, but speaking of certain values as if they are

priceless might also facilitate the formulation of policies that are cognizant of that

high valuation. It is important however to be cautious of dismissing intrinsic value-

speech altogether one the one hand, and of failing to realize that the use of such

language should not exclude the consideration of other valued ends, on the other

hand. Otherwise we run the risk of a different kind of alienation—becoming unaware

of other important considerations that we might be trading off.

177

Part I I I Rational Choice and Individualism

5 Liberal Respect for Identity? Only for Particular Ones

5.1 Introduction

In recent years, liberalism has been charged with failing to appropriately value

identity-groups based on social traits such as race, culture and religion. This charge

about the failure to recognize social identity can also be applied to rational choice

theory. As discussed in my Introduction, while rational choice theory only explicitly

endorses methodological individualism, a large area of rational choice theory at least

implicitly is committed to the stronger thesis of normative individualism, or what

Pettit calls ‘personalism’, that is prominent in the liberal tradition.1 Normative

individualism, as the name implies, is a doctrine about what constitutes the good. It

maintains that whatever is of value about nations, cultures, businesses, families or any

other group, is something that is valuable for the individuals they affect. In short, the

thesis denies that there are irreducibly collective goods.2 Thus, a large part of rational

choice theory—that which is committed to the liberal thesis of normative

individualism—would deny that there is value in social identity if identity is

1 This is especially true of welfare economics. Welfare economics is the branch of economics that is concerned with evaluating different institutions, actions, and policies, on the basis of normative judgments. The single most important analytical tool in this area has been the condition of pareto efficiency, which maintains that a state of the world, X, is morally better than another state of the world, Y, if in X at least one person is better-off and no one is worse-off than in Y. The pareto condition is an individualist account of the good since it only judges the desirability of these institutions, actions, and policies based on how well-off they leave individual persons. For a fuller explanation of these points, see Brennan (pp. 125-6) on the pareto condition and Pettit (pp. 22-30) on personalism or normative individualism. 2 Goodin argues that the claim that some goods (for instance, cultures, norms and language) are only possible in a society—a non controversial claim—does not show us that once these goods exist their value is not the value for individual persons but rather emergences from some new property of the group or collective itself. (Goodin, 1990, pp. 71-6)

178

described as only being a property of a group or collective itself. In this chapter, I

argue that if liberalism is to value identity groups, it will be constrained by certain

liberal principles. Even though these principles are not necessarily entailed by rational

choice theory, I will show how satisfying these principles will only support valuing a

form of social identity that is individualist. Thus, even though the explicit starting

point of the chapter is liberalism, the result is to countenance a form of social identity

that is also fully consistent with rational choice theory.

In what might be the most effective argument from the standpoint of

liberalism, communitarian critics argue that social identities are vital to the autonomy

and constitution of the persons who are members of identity groups and so the failure

to grant recognition and rights to identity-groups amounts to a denial of equal respect

for those persons.3 Any authentically liberal view that attempts to account for this

charge would have to be subject to certain constraints. Namely, in recognizing some

identity-group the account would have to not merely warn against harming the

autonomy of individuals in ways that have traditionally mattered to liberals, but

would have to specify a procedure to extend rights and privileges to identity-groups

that does not itself undermine the liberal sense of autonomy. With this constraint in

place, what would a liberal account of respect for social identity look like, and what

exactly would such an account be committed to respecting?

3 I am not referring to metaphysical discussions about identity or personal identity. Identity refers to traits that are shared by some people and not others, unlike the universal features of personhood, whatever we hold them to be. Another distinction is between personal identities—those that are not socially salient and defined, such as an odd sense of humor—and social identities— such as one’s race, ethnicity, culture, sexual orientation, religion, or disability. The arguments in this chapter do not hinge on there being a sharp distinction, although it is generally the latter kind that is central to concerns over respect.

179

In this chapter, I explore an ‘inclusive’ principle of equal respect for persons

in which we would respect both universal personhood and social identity. The

purpose here is not to argue that we have an obligation to recognize any social

identity, but rather to explore what implications would follow from the position that

liberals must respect identity as part of the respecting of persons. It will be shown

that even under this inclusive formulation of respect, we would need to distinguish an

identity belonging to particular persons from the general and collective forms of that

identity—for instance it would be Mike’s deafness and not deafness in general and

Nina’s culture and not the collective form of the culture that she shares with others

that would be the appropriate objects of respect. I will show how the need to preserve

the liberal sense of autonomy, together with the fact that people have multiple

identities at the same time, supports the distinction between particularized and

collective identity.

To determine which acts would be about respecting particularized identity

versus other identity forms, I provide a counterfactual test. The result is that any

action that passes this test would not seem to benefit a given identity at the expense of

other social identities an individual may have and it would not compromise the rights

and liberties of individual group-members. The account would thus satisfy the

autonomy constraint on an authentically liberal view of respect for identity. A major

practical implication of the counterfactual test is that collective rights would not be

endorsed. Because social identity is almost always viewed in terms of group-

membership, valuing identity seems to be at odds with the principles of liberalism,

democratic theory, and rational choice theory where persons are the ultimate source

180

of value.4 I will show, however, that respecting social identity would not necessarily

ascribe any special status to groups as such.

The arguments I offer regarding the necessity of drawing a distinction

between particularized identity and especially the collective form of the identity do

not deny the interdependent relations among persons, that people relate to and

genuinely care about others, or, especially, that identities are constructed in a

dialogical process in which they depend on others and are influenced by shared

institutions. The distinction I make is not between personal identities and social ones

but between particularized identities and collective ones. There is a consistently

liberal way to recognize social identity once we make this latter distinction. At the

same time, however, it is a central theme of this chapter to show that even if

communitarians are successful in their arguments concerning the value of social

identity to persons and their autonomy, there are important limitations on the

implications of these arguments.

5.2 Identity and Autonomy

There is a large literature on the psychology and value of social identity in

liberal democratic societies.5 The liberal conception of respect for persons—based

4 It is true that for Kymlicka in particular, the reason we must respect identity-groups such as cultures is because membership in identity-groups provides a context of choice for the individual—thus, cultures for instance are important ultimately because individual choice and autonomy matter. Members of transitional or waning cultures, he argues, are disadvantaged as compared to the majority cultures with respect to a stable context of choice, and so must be accorded special rights. However, Kymlicka’s view would also advocate special rights for the groups as such. 5 For recent significant works, see Appiah, Anthony (2005) The Ethics of Identity. Princeton: Princeton University Press; Benhabib, Seyla, ed. (1996) Democracy and Difference: Contesting the Boundaries of the Political. Princeton: Princeton University Press; Gutmann, Amy. (2003) Identity in Democracy. Princeton: Princeton University Press; Kymlicka, Will (1989) Liberalism, Community,

181

on the broadly Kantian and Rawlsian idea that persons are autonomous, self-

governing agents—has come under strong criticism by communitarians who argue

that the emphasis on a universal, rational human nature conflicts with our actual

psychology.6

Writers such as Michael Sandel and Charles Taylor argue that most of us view

ourselves and our choices, whether or not we want to, in terms of our (non-universal)

social identities such as race, ethnicity, gender, disability and culture, and they argue

that these identities are fundamental to who we are as persons because they are

fundamental to our capacity to value and choose—to our autonomy—the very thing

that liberals value so much. These critics argue that it does not matter that one’s

membership in some identity-group is often not chosen and is not reflectively or

consciously regarded as significant. In fact, that our social identities are not always

available for scrutiny—that they are outside our ability to reject or accept them—is a

key part of why they are viewed as being constitutive of who we are.7 They

constitute who we are in large part because they are the fixed frameworks in which

and Culture. Oxford: Clarendon Press, and (1995) Multicultural Citizenship: A Liberal Theory of Minority Rights. Oxford: Clarendon Press; Raz, Joseph (2001) Value, Respect, and Attachment. Cambridge: Cambridge University Press; Taylor, Charles (1989) Sources of the Self: The Making of the Modern Identity. Cambridge, MA: Harvard University Press; Young, Iris Marion. 1990. Justice and the Politics of Difference. Princeton: Princeton University Press.

6 Kant, Immanuel (1785/1983) Grounding for the Metaphysics of Morals. in I. Kant Ethical Philosophy Trans. James W. Ellington,. Indianapolis, IA: Hackett Publishing Co, and (1797/1999) Metaphysical Elements of Justice. John Ladd, ed. Indianapolis, IN: Hackett; Mill, John Stuart (1859/1975) On Liberty. David Spitz, ed. New York: Norton; Nozick, Robert (1974) Anarchy, State, and Utopia, New York: Basic Books; Rawls, John (1971) A Theory of Justice. Revised edition (1999) Cambridge, MA: Harvard University Press, and (1993) Political Liberalism. New York: Columbia University Press.

7Sandel, Michael J. (1982) Liberalism and the Limits of Justice. Cambridge: Cambridge University Press, 2nd edition 1999; Taylor, Charles. (1991) The Ethics of Authenticity. Cambridge, MA: Harvard University Press.

182

we figure out our values and goals, and this is especially true for people whose social

identities are non-voluntary, such as those living in traditional societies and those

whose identities have a physical component, such as being deaf or having a particular

race.

Taking a slightly weaker position, Will Kymlicka describes social identities,

especially cultures, as providing background contexts that give our choices and

actions meaning.8 One’s culture is a largely un-chosen and deeply embedded part of

who one is and forms the basis from which one determines a conception of the good

and, in particular, within certain boundaries of what is considered appropriate, good

or valuable.9 What all of this means, according to the above critics, is that if the

concept of a person and of a person’s autonomy essentially includes social identity,

then the liberal principle of respect for persons must also include this identity.

While the notion of respect can mean many different things, what it means in

this context is to accord certain rights and privileges to identity-groups that have been

faced with disadvantage. For instance, it might involve exemptions from public

education requirements for traditional societies, financial support for a private school

for deaf students, or providing resources for the construction of a mosque in order for

8Kymlicka in Liberalism, Community and Culture and Multicultural Citizenship 9 There are reasons for believing that certain identities would be more significant to autonomy than others: those that have been met with social obstacles, whether disadvantage, discrimination or both. It seems these identities would be especially salient for one’s outlook and understanding of the social and political world, and therefore especially important to one’s capacity to value. This is because while disadvantaged identities limit some opportunities, they create others. In fact, the idea that identities serve as constraints seems to a large extent to be the reason they are also viewed as providing options. Opportunities, at least meaningful ones, do not appear in a vacuum. They are generated and gain their meaning in relation to certain restrictions. In this way, disadvantaged identities would seem to provide an even more significant “context of choice” than other identities. (See Kymlicka, Multicultural Citizenship, pp. 82-84. Kymlicka uses this term to refer to all sorts of cultures and not just those that have been disadvantaged)

183

a Muslim community to feel that they are fully accepted in a western society—

something that, advocates would argue, would not be seen as establishing any religion

but rather recognizing one that is at a disadvantage. Importantly in all such cases, the

special benefits that the group receives would be seen as a way to accord equal

respect to certain social identities that are at a disadvantage either because they are

not part of the majority social identity or because they have been discriminated

against or both. Thus, the benefits would not be viewed as in fact being special.10

Even if critics are correct that liberals must respect social identity in these

ways, there are certain constraints internal to the liberal tradition that any liberal

account would be subject to. A liberal account could not include those social

identities that harm others or, more generally, that are not tolerant of others. If some

group, for example, believes that all non-group members are morally inferior, liberals

would have reason not to recognize that group. Many authors acknowledge that

respect should only be extended to identity-groups that are tolerant of other groups.11

This can be called the ‘tolerance constraint’. Of course understanding precisely what

it means to be tolerant of others is a contentious matter in liberal and democratic

theory. Putting this difficult issue aside for now, it is not just the harm that may come

to other groups that must be guarded against.12

10 Kymlicka, Multicultural Citizenship, see especially Chapter 5 11Gutmann, Identity, see especially pp 89-112. Liberals would not be committed to respecting morally repugnant identities, such as a deep hatred for other races, no matter how integral to one’s sense of self. If one’s identity is only partially defined by such an element, however, we may be able to respect other elements of the identity without embracing this aspect. 12 There is also reason to believe, as Gutmann argues, that groups that do not respect the autonomy of individual members will likewise not respect other groups.

184

More important for the goal at hand, it is a minimum constraint on any liberal

account of respect that the autonomy of the members of the identity-group must not

be harmed in ways that matter to liberals.13 In other words, even if one were to agree

with communitarian kinds of arguments concerning the importance of social identity

to autonomy, a liberal account of respect for identity would need to ensure that the

liberal sense of autonomy would not be undermined. As Kymlicka has pointed out,

liberals are committed to protecting the freedom of individual members of an

identity-group to question and revise the practices of the group and, more specifically,

to protecting their basic rights and liberties.14 We can call this the ‘liberal autonomy

constraint’. It would not be a liberal account if autonomy could be enhanced in ways

that would address the communitarian arguments but would be undermined in ways

that are important to the liberal tradition. The latter is a necessary condition for any

liberal account to be consistent and is what would set it apart from a communitarian

view.

But I want to emphasize that a liberal account must do more than merely

caution or assert that respecting social identity should not harm those aspects of

autonomy important to the liberal tradition, as other leading liberal accounts have

only done.15 It must be able to specify a way to recognize social identity that does not

13 As Kymlicka has pointed out, “one can ensure tolerance between groups without protecting tolerance of individual dissent within each group”, p. 162 14 Kymlicka, see especially pp. 152-158. 15 Kymlicka eloquently argues that liberals can endorse ‘external protections’--protecting the resources, institutions, and practices of the group against outsiders--but not ‘internal restrictions’-limitations on the liberty of some group-members imposed by other members. The latter would be a violation of the basic rights and liberties of individuals. However, a problem arises because he does not offer any procedure or guidelines for determining whether a particular practice would fall under the label of external protections or internal restrictions. For instance, is the imposition of marriage on young women (or the demand to exempt children from public education requirements) in a particular

185

itself undermine the liberal sense of autonomy. More to the point, there must be a

built-in safeguard against violating the basic rights and liberties of individual group

members. Otherwise, the account would fail to offer any assurances that the

autonomy of the individuals who are the object of such identity concerns would be

enhanced overall rather than diminished, and thus fail to satisfy the expressed aim of

the account.

The overwhelming focus in the debate on social identity has been on

membership, whether formally or merely associative, in identity-groups that are

marked by a shared trait or characteristic. The question of whether these groups

ought to be able to collectively pursue their individual interests or collectively

exercise their individual rights has not been an especially contentious matter in liberal

discourse.16 Much more contentious is whether an individual’s rights, status, or

obligations as a member of an identity-group should ever trump an individual’s

rights, status, or obligations as a person. Because social identity is almost always

group an internal restriction, because it restricts the liberties of young women (or children) by other group-members, or is it an external protection, because it shields the institution of imposed marriage (or the community’s norms about education) against the political powers of the majority? Since Kymlicka endorses not just territorial autonomy and language rights as significant external protections, but also a group’s veto powers over the larger society’s decisions that are of importance to the group and self-government rights, it is not clear where the examples just given, and many others like them, would be categorized. (see especially pp.34-47 and 108-115) The account I offer is not necessarily at odds with Kymlicka’s. It may offer a justified procedure for determining which external protections should be endorsed.

Likewise, in a very recent article, Wall argues that there is autonomy-based support for collective rights and appropriately cautions that such rights could not infringe on the basic liberties of individuals in order for them to be consistent with a distinctively liberal account. However, he does not show us how these rights would not in fact undermine individuals’ liberties and offers no specification for which collective rights would and would not violate the liberal conception of autonomy. (Wall, Steven “Collective Rights and Individual Autonomy,” 2007, Ethics 117, n2, pp 234-264.) 16 As Appiah notes, these collective rights “tend to have more friends…Most people think that it is just fine that Utah or the city of Cambridge or the Catholic church can exercise rights, through the ballot box or (in the case of churches) through whatever consensual internal mechanisms they agree upon.” p. 72-3

186

viewed in terms of group-membership, valuing identity seems to be at odds with the

principles of liberalism and democratic theory where persons are the ultimate source

of value. 17 I will show, however, that respecting social identity in such a way that the

autonomy constraint is satisfied would not ascribe any special status to groups as

such.

Before demonstrating this, I must first discuss what it would mean to respect

social identity as part of the respecting of persons. To sum up the communitarian

position, social identity is considered to be significant to who one really is. As such,

it is held to be an important source of claims on liberal societies. Bhikhu Parekh puts

the claim succinctly when he writes that ‘the liberal is in theory committed to equal

respect for persons. Since human beings are culturally embedded, respect for them

entails respect for their cultures and ways of life.”18

5.3 Including Identity in Respect for Persons

So what would it mean to respect a culture, race, or “way of life” because it is

important to who someone is, to her autonomy? As suggested, it would mean

including these identities as part of the respecting of persons. Under such an

17 It is true that for Kymlicka in particular, the reason we must respect identity-groups such as cultures is because membership in identity-groups provides a context of choice for the individual—thus, cultures for instance are important ultimately because individual choice and autonomy matter. Members of transitional or waning cultures, he argues, are disadvantaged as compared to the majority cultures with respect to a stable context of choice, and so must be accorded special rights. However, Kymlicka’s view would also advocate special rights for the groups as such. 18 This is the view of several prominent political theorists. For instance, writing in regard to culture, Kymlicka argues that the liberal principle of equal respect for persons requires the recognition of the rights of certain cultures since culture is considered “a constitutive part of who the person is” (Multicultural Citizenship p. 175) and Taylor argues likewise that an important condition of respecting someone is recognizing that person as being constituted by her culture. (1992) Multiculturalism and the Politics of Recognition. Princeton: Princeton University Press)

187

‘Inclusive Principle of Respect for Persons’, or Inclusive Principle for short,

individuals who are persons would be respected for their primary personhood, where

personhood includes the liberal sense of autonomy that all persons have in common.

In addition, in order to fully respect persons, we would also, in the same basic way in

which personhood is respected, respect social identity. This is the position that

communitarian writers and advocates of minority rights are committed to adopt when

they argue that by failing to recognize the identities of certain groups of people,

liberals disrespect those people as persons; not only their identities and, especially,

not merely in some way.19 Since the claim is that a social identity is inseparable from

who a person is and therefore is to be respected in the same way in which personhood

is, instead of considering respect for identities as a separate moral demand, we can

view it as a different mode of respect for persons. The two different modes of the

Inclusive Principle are:

(1st) Persons must be respected for their personhood (2nd) Persons must be respected for their social identities

I must immediately state what I am not attempting to do here. First, I am not

seeking to provide the conditions for picking out the social identities that would be

included.20 It does seem plausible to include only certain categories of social

19 Sometimes when identity-groups, or those that advocate on behalf of them, argue that they are being denied equality or respect, they do not explicitly use the terminology of ‘equal respect as persons’. But it seems clear that their objection is at bottom about being denied equal respect as persons, and again not that they are being disrespected merely in some way. The real concern generally seems to be that members of some identity-group are being deprived of a basic or fundamental respect—that of respect for persons. 20 For instance, it seems that there is a first-person authority concerning what identities are constitutive, but that there must also be some public epistemic constraints, such as that the individual must actually possess the identity as seen third-personally. Questions such as who determines whether an identity is constitutive (and furthermore, meaningfully constitutive) are not explored here.

188

identities, such as again those that have been met with discrimination or disadvantage

and those that do not harm others, but this issue is beyond the scope of the chapter.

Second, and more importantly, even once we have such conditions in place for which

identities qualify, I am not arguing that the value of social identity is of equal weight

with the value of persons for their primary personhood, even if the identities that

qualify should all be valued equally with one another.21 Since this is a liberal

account, it endorses the traditional liberal assumption that for an individual who is a

person, her primary moral status is her personhood, universally defined. Social

identity is derivative of and dependent on personhood in at least two ways. In the

first place, we could not even say that social identity is central to who we are as

persons were it not for the personhood features already in place, so to speak.

Personhood is constitutively or causally primary—being someone who values and

chooses at all is prior to being someone who values and chooses in the particular way

that communitarians maintain. More importantly, under the kind of communitarian

charge presented, the reason that liberals would be committed to recognizing social

identities is, again, because of the persons they comprise.

The primacy of personhood is the basis for the tolerance constraint.

Respecting some identity-group cannot violate the 1st mode of respecting persons by

harming non-group members or by depriving them of their basic rights and liberties.

21 Although there might be important distinctions between respecting someone and valuing someone, I will assume that both notions refer to a basic sense of respect that is involved with the notion of respecting persons--which might be called recognition--and not a more robust or demanding sense of valuing, such as that which might be involved with esteem, admiration, or honor. Furthermore, in some formulations of respect, respect includes the notion of valuing. Respecting any object simply means that the object must be valued in an appropriate way, and not promoted or used. For instance, see Philip Pettit’s discussion in his “Consequentialism and Respect for Persons,” 1989. Ethics 100, pp 116-126.

189

However, there is a trickier, and for the goals at issue in this chapter, more critical

way in which the 1st mode of respecting persons must not be violated; as discussed

before, respecting some identity-group cannot deprive the individual members of that

group of their basic rights and liberties. In the next section, I will begin to

demonstrate how the account here, by maintaining distinctions in identity form,

satisfies the liberal autonomy constraint.

5.4 The Locality of Respect

While the Inclusive Principle does not imply that social identity is as valuable

as personhood itself, it does suggest that they are valued in the same basic way

because both social identity and personhood are important features of persons. This

point brings out an important aspect of respect for persons that must now be made

explicit: only traits or features of persons would be included. This might seem a

trivial point. Of course something like the Grand Canyon would not be included

because the principle is about respect for persons. However, in order to understand

more about the form of identity that would be included, it is helpful to express why

other objects would be excluded. Importantly, it would not be due to the worth or

significance of the object to us or to individuals’ lives. The Grand Canyon may have

extraordinary value. Instead, the reason that it would not be included in respect for

persons is that, simply put, it is not a trait or feature of individuals who are persons.22

22 Not only would something like the Grand Canyon be excluded from the respecting of persons, but it would also seem to be excluded from having the type of value that persons have. This separate claim is not strictly contained in the idea of respect for persons—we could after all hold that both persons and the Grand Canyon have the same kind of worth —but it seems to be related in an important way. In order for the idea that persons are equally worthy of respect to be of significance, it seems only persons

190

Respect for persons does not directly include respect for anything other than persons,

and persons are their traits and features. This does not mean that all traits and features

of persons would count for respect, but rather that only these would count.23

This claim applies straightforwardly to social identities that we deem

valuable, even highly valuable, but which are not features of persons. Claiming

something to be valuable or important to persons is different than claiming something

to be an important part of persons.24 This is in fact a primary reason that social

identity takes on the significance it does: identity is not just important to persons. In

popular and scholarly discussions, social identities are seen as being an important part

of persons. However, there is a crucial qualification that must be made to the last

point: an identity must belong to individuals who are persons in order to be

considered a trait or feature of persons and thus to be included in respect for persons.

While it may seem obvious that only identities belonging to persons would be

included, there are distinctions in form that identity can take. What I am contrasting

is particularized identities from more general types of identities. To say that

something ‘belongs to’ or is a ‘part of’ persons is to say that it belongs to or is a part

of particular persons. Identities in general or collective identities on the other hand, must be valuable in the way that they are. The force of the principle comes in large part from the singling out of persons as opposed to other objects of value. Even though other objects would not be included in the principle, this does not imply of course that they do not have some other significant value. To take just one example, many non-human animals that do not qualify as persons still have an important value and should not be subject to cruelty and inhumane treatment. Furthermore, this does not imply that the value of persons is incomparable with the value of non-persons or other things. In fact, to say that the Grand Canyon is not of equal value with persons is to compare the its value to the value of persons. 23 Dave Estlund gave me a nice example here: personhood respect does not entail fingernail respect. 24 These concepts do come apart. Art, family, and a deep passion for traveling can be very valuable to us without being part of anyone’s identity. At the same time, we can have features of our identity that are trivial or even shameful. While we can consistently and perpetually be ashamed of something that is a part of us, it seems the same does not hold for something we value; we would come to no longer value it.

191

do not belong to anyone in particular. So, for example, if Mike is deaf, Mike’s

deafness or Mike’s being deaf would count but deafness in general or deafness as

common to members of the deaf community would not.25 Only particularized

identity—Mike’s deafness—is localized enough to persons to count in respect for

persons.

I believe that there is intuitive support for deconstructing identity in this way,

as I will try to show here and in the next section. However, even if one were tempted

to deny these distinctions, I will discuss how maintaining them, and taking actions to

benefit identity accordingly, provides a means of recognizing identity that satisfies

the liberal autonomy constraint. While it may not be the only conceivable account of

respecting identity that would satisfy this constraint, I will show that the account

offered here fully satisfies the autonomy constraint, genuinely advances identity-

based interests, and does not make, I believe, implausible distinctions in identity.

Thus there are several, independent, reasons to endorse these distinctions.

The distinction between a particularized identity, on the one hand, and the

collective or general form of the identity, on the other, proceeds from two related

claims. The first, and mainly normative, claim is that one’s attitude towards one’s

identity is partly constitutive of that identity. The second, and descriptive, claim is

that a given identity belonging to an individual will interact with the other identities

that she has. Taking the first claim, it seems a liberal account of the importance of

identity should want to maintain that someone’s identity will depend not only on the

25 There may be important distinctions that must be made between a collective identity and an identity in the abstract, but these distinctions are not necessary for my central purpose here of distinguishing the identities belonging to persons from the less particularistic forms.

192

ascription of an identity to her but also on her self-conception of that identity. We

can recognize that whether one has a certain culture, race or ethnicity is largely

beyond one’s control and, because of this involuntariness, will be important to

autonomy—this is the communitarian point. However, this does not commit us to the

view that how one conceives of that identity is beyond one’s control. Indeed, any

liberal autonomy-based account should resist the latter view in order to ensure that

the liberal sense of autonomy is preserved. It is important that any liberal account of

identity place emphasis on an individual’s own formulation, revision, and valuation

of an identity. Maintaining that one’s attitude towards one’s identity is partly

constitutive of that identity achieves this emphasis.

Talk of the importance of identity typically takes on the language of how an

individual belongs to an ethnicity or culture or race. But there is also a significant

sense of belonging, captured in the liberal view of autonomy, which goes in the other

direction. There is a sense in which an ethnicity, culture or race belongs to an

individual because of that individual’s distinctive conceptualization of the identity.

This is supported by the second (descriptive) claim above—the fact that people have

multiple identities at the same time. Although someone’s given identity may be

considerably defined in a way that also defines the relevant identity-group, it would

not seem to comprehensively define her identity. There will be other identities that

are important to her conception of a given identity, whether these other identities are

shared with other people or highly personal. Different identities might range in their

significance to the autonomy and composition of a person, and communitarians may

be correct that one’s culture would be especially significant, perhaps much more so

193

than one’s gender, for instance. However, the fact of multiple identities means that a

given identity belonging to an individual takes on a unique character because it is

combined and integrated with that individual’s other identities, even if the identity at

issue is a defining or central one of the individual.

Amartya Sen has said that he is, at the same time:

“an Asian, an Indian citizen… an American or British resident, an economist, a dabbler in philosophy, an author… a strong believer in secularism and democracy, a man, a feminist, a hetero-sexual, a defender of gay and lesbian rights, with a nonreligious lifestyle, from a Hindu background, a non-Brahmin, and a nonbeliever in an afterlife (and also, in case the question is asked, a non-believer in a “before-life” as well).”26

Importantly, this sort of amalgam of identities seems to apply to all of us. It is

not only a feature of modern societies and the mixing of identities they involve. A

mother in the traditional village of Kasur, for instance, is, at the same time, a woman,

a parent, a mother, a Muslim, a Punjabi speaker, a member of a particular family, of

the province of Punjab, a Sunni, a Kasuri, and a Pakistani. Of course, this description

only captures (some of) her public identities. Were we to consider her personal

beliefs, preferences, habits, and lifestyle, we could say much more. Even if it were

true that being a woman is not as significant to who she is as is being a Pakistani, the

fact that she is also the former means that her experiences and conception of what it is

to be Pakistani will most likely not match those of a man who lives in Pakistan, and

the same point can be made concerning the effect of her other identities on her

identity as a Pakistani. Furthermore, the assorted identities someone has will vary in

their salience depending on the context. In a conflict with India, being a Pakistani 26 Sen, Amartya. (2006) Identity and Violence: The Illusion of Destiny. New York: W. W. Norton & Company. p. 19

194

might take prominence, but in the matter of “honor killings” on the sub-continent

being a woman might take priority over being a Pakistani. So different people with

the same identity can prioritize, value, or conceive of that identity differently because

of the other identities they have. Drawing the distinctions in identity form in the way

that I have fully allows for these differences between individuals’ attitudes towards

their identities.

To better understand the distinctions in identity form and why only

particularized identities would be included in respect for persons, consider an

example. Nina’s South Asian heritage is integral to who she is. She has grown up to

think of herself as South Asian and her heritage provides a broad context for choices

that she makes throughout her life and has shaped her values and goals. However, it is

the fact that it is Nina’s heritage, that makes it integral to her identity as a person, to

her autonomy, and not that it is just any South Asian heritage or a South Asian

heritage more generally. Notice that we would not consider Aysha’s South Asian

heritage to be an important part of Nina. Nina’s attitude towards her heritage will

almost certainly be different than Aysha’s attitude towards hers. The fact that each

person has several other identities at the same time means that the heritage of each

takes on a particularized character because of its interaction with those other

identities. Furthermore, each person’s conception of what it means to be South-Asian

is likely to be shaped by how relevant and salient being a South-Asian has been in

terms of her history and development. Thus, the identity belonging to Nina is

distinctive to her self-conception because of how her identity has figured in her

history and development and the way in which it interacts with her many other

195

identities. All of this changes the nature of the identity. It particularizes it and makes

it a part of who Nina is as a person in some basic sense.

This is not to deny that a South Asian heritage in general can be very

valuable to Nina in a number of ways. The forms of identity range in the role they

play in a person’s practical orientation and sense of self. Nina might think it very

important that there continue to be people from South Asia and she might value the

art, food, music and languages of that region. And mutual identification with others

who are also South Asian may be a sufficiently strong form of identity to share

certain ends and values, and also to use their shared identity to secure political ends

such as obtaining greater resources and ensuring fair and equal treatment. Thus even

the more collective form, ‘our South Asian-ness’, is important to Nina. But for the

purpose of including social identity as part of persons, the collective form of South

Asian culture would not seem to be sufficient. It is not that the importance Nina

places on the collective identity is low, but rather that the collective identity is

insufficiently localized to constitute who she is.27

The distinctions in identity form seem even stronger if we turn from culturally

or socially conceived identities to physical or biologically rooted ones, because the

latter are in a sense located in or attached to individuals and so are even more prone

to being uniquely viewed and thus even more localized. If Mike is deaf, it would be

his deafness and not the deafness of another person that would seem to be significant

to who he and to his autonomy. We would not say that the type of deafness that Mike 27 More accurately we might say that Nina’s having of a South Asian heritage would be unique or particular to Nina. It is not the identity per se that changes when it is instantiated in an individual, but rather there is a particular way that individual possesses or embodies the identity. I am grateful to Jerry Levinson for clarifying this point to me.

196

has but present in another person, Tim, is formative of Mike. Nor would deafness in

general be identity-forming of Mike—after all, what seems to be crucial in observing

that Tim’s deafness would not be a part of Mike is not that the deafness belongs to

Tim, but rather that the deafness does not belong to Mike. Mike will have his own

experiences of what it means to be deaf, even if it is largely shared with others who

are also deaf.

The preceding discussion does not suggest that a person’s social identity must

comprise the whole of an individual in order to be particularized and to qualify for

being recognized in the principle of respect. On the contrary, as discussed above,

people have not one but many identities at any given time; some of them social and

some of them highly personal. This is an important psychological fact about persons

and their identities that must be accounted for in any account of respect for identity,

and especially any liberal one. A social identity takes on the particularized character

it does when it belongs to a person in large part because it is combined with, and

stands in a particular relation to, other identities of that person. A given identity will

only be one feature of a person, possibly a central or significant one, but must be one

that is particularized to a person if it is to count here. Social identity that is not

particularized to persons would be excluded from the respecting of persons in a

similar way that the Grand Canyon would be. For while it might be something we

deeply value, it would not constitute a particular person.

197

5.5 Collective vs. Particularized Identit ies

Many of the social identities that are central in the debate and politics of

respect for identity may seem to complicate the contrast that I have drawn between

particularized and collective identity. Culture is the most obvious example. Someone

might argue that the collective culture is her identity. Such a person would not need

to maintain that the collective culture makes-up the whole of who she is—only that

the collective culture is her culture--that part of her whole person that is made up by

her culture. So, this objector claims, the only way for her cultural identity to be

respected is to respect the collective cultural identity. Rather, the collective identity

cannot even be meaningfully distinguished from the particularized identity, at least in

terms of applying the principle of respect for persons.

In order to be included in respect for persons, it should be emphasized that the

collective identity would indeed need to be equivalent to the particular identities of

persons who are members of that identity-group.28 In other words, it is not sufficient

to maintain merely that there are overlaps and features in common between the

collective culture and the culture belonging to Nina. This weaker position would be

perfectly compatible with the distinction I have drawn. In order to collapse the

distinction then, one would need to say that the collective cultural identity is identical

to the cultural identity of particular persons. However, there are problems with this

equivalence.

28 Again, one would not need to maintain that the collective identity in question is equivalent to the whole identities of particular persons, only that the given collective identity is equivalent to the counterpart identity in those persons.

198

One way to summarize the normative claim made above regarding the liberal

sense of autonomy is to caution, as Anthony Appiah has, that we could be trading one

tyranny for another—that of the larger society for that of the group.29 Emphasizing

the collective nature of some social identity may liberate the identity-group from the

norms of the larger society, but can also unduly subject individual members of the

group to the norms of the group, and in particular, to how the group construes their

shared identity. Even apart from this distinctively liberal response, though, it seems

collective and particularized identities can and do come apart, not only conceptually

but also practically, as I will briefly try to show here and will discuss at greater length

in the next section.

In popular and political discussions, people often point out that disrespecting

someone who is like them, but not them specifically, makes them feel that their own

identity is being disrespected. However, while it is common to feel if one disrespects

someone who is like me, but not me specifically, that my own identity is implicated,

this seems less true in scenarios in which one respects or benefits someone like me

but not me specifically. If Aysha’s boss were to grant her special time off for

celebrating a South Asian holiday, this would unlikely make Nina, who works at a

different firm, feel as if the South Asian heritage belonging to her was being

benefited, at least not to the extent that it was for Aysha. Even if Nina were to know 29Appiah eloquently warns “the politics of recognition, if pursued with excessive zeal, can seem to require that one’s skin color, one’s sexual body, should be politically acknowledged in ways that make it hard for those who want to treat their skin and their sexual body as personal dimensions of the self. And personal, here does not mean secret or (per impossible) wholly unscripted or innocent of social meanings; it means, rather, something that is not too tightly scripted, not too resistant to our individual vagaries.” (original emphasis) Appiah, p.110. However, since he does not explicitly consider the distinction I make between a general type of identity and an identity belonging to persons, he worries that any view that supports recognition for one’s identity is susceptible to the charge of social imposition.

199

that she would also get time off if she were working for Aysha’s boss, it would still

be the case that Nina’s South Asian heritage would not currently be benefited in the

same way or to the extent that it was for Aysha.

It could be objected that I have only shown that Nina’s whole person is not

being benefited to the extent that Aysha’s is, but that I have not shown that Nina’s

cultural identity is not receiving the same benefit. In other words, it could be said that

what I have demonstrated is that it is the other parts of Nina, other than her culture,

that are not being benefited, and this is why there is a difference in the way that

Aysha and Nina experience the action of Aysha’s boss. But it seems that if this were

true, then a good case could be made that one’s culture is not a very significant part of

one’s whole person. For if the other parts of Nina are such that they are capable of

diminishing or countervailing the presumed beneficial effect on her cultural identity

to the point that Nina as a whole does not benefit, then this serves to undermine the

claim that culture is basic to who Nina is as a person. On the other hand, my claim

that Nina would not benefit to the extent that Aysha would does not rest on the idea

that other parts of Nina would counteract the effects on the part of her that is her

culture. My claim is that Nina’s culture would not benefit to the extent that Aysha’s

would and this is why Nina would experience the action of Aysha’s boss differently

than Aysha would. This claim does not undermine the point that Nina’s culture is

basic to who she is as a person and in fact supports it.

There are of course many examples of actions that would equally express

respect for both the particularized and collective forms of the identity at the same

time, and I will say more about this in the next section. But the point here is that

200

there also seem to be examples of actions that would express respect for one person’s

identity without doing so, or doing so to a lesser degree, for another person’s

identity—and this is what is needed to establish that particularized and collective

identities come apart. To take the analogy with personhood features, suppose that

Bob’s capacity to value was being respected in some very generic and minimal way.

His neighbor tells him that he admires the bare fact that Bob values things. Even

without referring to some special or intense way that Bob values, his neighbor’s

comments would unlikely demonstrate respect for Carol’s capacity to value in the

same way or measure as it does for Bob’s. And it seems social identities are more

capable of being unique to the individuals who possess them than are personhood

features. Social identities are richer than personhood features and therefore allow

greater room for variation, hybridization, and interpretation across individuals. While

the features of personhood are common enough to apply universally to all persons,

social identities of course are not.

It might seem that I have only illustrated the gap between an identity

belonging to one person from the identity belonging to another person, and not

between a particularized identity and a collective one. This would be an important

objection because the point of saying that one’s culture is one’s identity is not to

stress that the identities of different members of an identity-group are equivalent, but

to maintain that the collective identity is indistinguishable from the identity of the

persons who belong to the group. However, the fact that one member’s identity is

different, even if only in small ways, from another member’s identity suggests that

there is a collective identity that stands separate from both.

201

If we say that the collective form of South-Asian culture is Nina’s cultural

identity, and Aysha is also a member of this culture, than we would have to say that

the collective identity is also Aysha’s identity. Otherwise, we would be maintaining

the odd view that while Nina’s identity is the collective one Aysha’s is not. Someone

who takes the hard-line view of claiming that a collective identity is indistinguishable

from the identity of someone who is a member of that group would not be

maintaining that only one member’s identity is the collective one, of course, but that

every member’s is the collective one. So the collective identity would be

indistinguishable from both members’ identities. If this were the case, however, it

would mean that we could not meaningfully distinguish between one member’s

identity and another member’s identity. But I have tried to show that we in fact can,

largely because of the role of multiple identities. If we can make this distinction, then

there is a collective form of the identity that is also separate from both these

members’ identities. The fact of multiple identities and the particularized fusions they

give rise to does not mean that we would need to respect each instantiation of an

identity in some unique fashion—that would obviously be very difficult to do. It does

suggest, however, that there is a gap between the identity of one person and that of

another, and thus between a collective and particularized one.

None of the preceding discussion denies that persons depend on others for the

construction of their identities or that people genuinely care about others. Political

theorists such as Sandel and Taylor are correct to argue that people value

relationships of mutual identification and support. Furthermore, the discussion has

not attempted to refute the claim that many of our identities are inherently social, in

202

both their causes and constituents, as it is these that are central in concerns over

identity. Nina’s South Asian heritage is of course social in nature and dependent on a

shared history—what would it even mean otherwise? But it would still only be the

social identity belonging to Nina that would be included in respect for persons. The

distinction I have drawn is between particularized identities and collective ones, not

between personal identities and social ones.

As I said above, I believe there is intuitive support for unpacking identity in

the way I that I have. However, even if one were tempted to deny these distinctions, I

have tried to show the liberal rationale for maintaining them. In the next section, I

will discuss how taking actions on behalf of particularized identity provides a means

of benefiting or promoting identity that satisfies the liberal autonomy constraint, yet

still genuinely recognizes identity.

5.6 Practical Implications of the Distinction

To restate the central argument, if liberals are committed to respecting social

identity as part of what it means to respect persons, the commitment would be limited

to recognizing social identities belonging to particular persons and not general or

collective ones. This is because it is only particularized identities that would be

constitutive of persons and their autonomy. I argued that there is a conceptual and

normative gap between particularized identities and, especially, collective ones, but

there is also a practical gap between them.

We can begin by asking what kinds of actions would benefit particularized

identities as opposed to other forms. It might seem that pragmatically speaking there

203

would be no difference; whatever action benefits social identities belonging to

particular persons does so because it promotes the broader category. However, there

is a useful counterfactual test for determining which actions would benefit

particularized identities without necessarily benefiting the collective or general form:

that of examining whether the action would benefit or promote an identity-based

interest if only one individual were affected by it or, more generally, if the action

would refer to only one instance of the identity in question. That this kind of query is

intelligible also provides us with another reason for making the conceptual distinction

between particularized and collective identity. Again, they do come apart in

important and meaningful ways.

It should be stated that I am not urging that we are required to take actions that

pass this test for respecting social identity, nor that all such actions would be cases

related to the concept of respect, as opposed to other moral or political demands.

Rather, my concern is in determining which actions would be justified in terms of

particularized identities. Proposed actions that pass the counterfactual test would

represent a prima facie identity-based reason for action. Whether we ought to take

such an action or implement it into policy would depend on a number of conditions,

such as practical feasibility, the efficacy of the action in promoting significant

identity-based interests, costs, and the need to avoid disrespecting the identities of

other groups. Furthermore, even though the test specifies that an action must be able

to refer to only one instance of the identity, the purpose of implementing the action

would of course be to benefit several people who possess that identity—to benefit an

identity-group.

204

To see how the test would operate, consider expanding the current policy

under the American with Disabilities Act of requiring wheelchair accessibility in

public buildings to pertain to all private buildings as well. Even if there were only

one person living who required a wheelchair, this policy would benefit the identity in

question. Although it may require a community of wheelchair users to warrant the

associated costs, and it would of course potentially benefit every member of this

community, the action would not require such a community in order for it to confer

respect to someone who depends on a wheelchair. In this way, we can see that the

policy would be justified in terms of the particularized identity. It may also affect the

collective identity of course, but even in the absence of others with the relevant type

of identity, the action would still promote an identity-based interest. To take another

example, a policy that allowed Muslims to take five breaks during the workday, for

the purpose of prayer, would benefit a single Muslim person. It would be a

meaningful expression of recognition even in the absence of other Muslims. Thus the

action would be about respecting a Muslim identity belonging to particular persons.

Certain acts of recognition for identity would only be meaningful in the

context of an identity-group. Language, which is an important element of many

cultural identities, presents a good example since it depends on a community of

speakers. Teaching a minority language in school (in addition to the majority

language) would really only be beneficial if there were more than one speaker of the

language in existence. In the absence of other speakers, there would be little point to

205

introducing it to anyone.30 However, there are some actions that would respect a

language, such as posting signs written in a minority language, which would benefit

the identity if only one person spoke the language, even though it is an implausible

counterfactual. None of these examples of actions taken on behalf of particularized

identity would be trivial. They each would promote significant identity-based

interests.

I said that the distinction I have drawn is between particularized identities and

collective ones, not between personal and social identities. The concerns raised over

identity almost always pertain to social identities such as culture, race and ethnicity,

and so the Inclusive Principle would pertain to social identities, albeit, not in their

collective form. However, since the counterfactual test calls for considering actions

that must be able to refer to only a single instance of an identity, it might be objected

that I have not shown how genuinely social identities, which necessarily involve

others, would be included. In response to this charge, we can note that while it is true

that we can only make sense of a social identity in the context of others, this does not

mean that we can only make sense of actions about a social identity in the context of

others.

For an action to respect a social identity, it must promote interests of a certain

kind: those that individuals have as a result of being part of a group or society. Under

the counterfactual test, an action must be able to benefit any single individual’s

interests, but this would include those interests that arise out of membership in an

identity-group. The concern to pray five times a day is something that comes from 30 Indeed teaching any language would only be possible in the context of more than one speaker of the language, as the teacher would already have to know the language.

206

taking part in the Islamic religion. Likewise, one’s interest in having signs written in

her minority language is one that is born from being a member of that language

community. In this way, both an allowance for breaks for prayer and special signs

would benefit social identities.

Furthermore, as I mentioned above, an action would need to do more than just

pass the counterfactual test if it has any chance of becoming policy. It would also

have to be worth incurring the costs of implementation. In most cases this will mean

that the action would potentially benefit a large community. What such an action

would do then is to pick out those features of a given identity that members of some

identity-group have in common and so would be about the shared features of that

identity. It would not promote the individualized features of identity or those that are

blended with someone’s other identities.

Finally, even though the justification of these actions would need to be in

terms of benefiting particularized identity, this does not mean that the policy would

not in practice also promote the collective form. Putting up signs in a minority

language would do more than benefit each individual speaker of that language. It

would also facilitate coordination and planning of shared activities of the group of

speakers of that language, and thereby benefit the collective identity as well. It is not

as if the action would have to benefit individuals alone or in private. To the contrary,

actions that would be justified in virtue of respecting particularized identity would

often enable members of an identity group to come together. Granting Muslims

prayer-breaks would benefit a single Muslim, but would also facilitate increased

assembly among Muslims that work in close proximity to each other—if they would

207

be able to take breaks during the workday for prayer, they would have a greater

opportunity for meeting with their Muslim co-workers than if no such breaks were

allowed, and this would in turn benefit the collective identity of a community of

Muslims.

At the same time, though, and this is an important point, collective rights

would not be endorsed.31 There are different categories of collective rights. In one

category are rights justified by reference to the interests of more than one individual

or to the aggregated interests of each individual to coordinate their activities. An

example of this might be preserving a minority language. In a second category are

rights that refer to a collective goal, often expressed through institutions of authority,

and not merely to the aggregated interests of individuals. Advocates of this second

category of collective rights claim that the group has a shared interest in maintaining

certain norms even if some individual members of the group do not endorse the

practice. Examples include the right of Muslims to enforce the wearing of

headscarves by women in their community and, as in the famous Wisconsin vs.

Yoder case, a right of elders in a society to exempt their children from government

education requirements.

31 Charles Taylor argues, with respect to culture, that in order to protect an individual’s identity we must try to preserve the distinctive tradition of the identity-group, which would imply support for collective rights of certain kinds. (1994) Multiculturalism: Examining the Politics of Recognition, ed. Amy Gutmann. Princeton: Princeton University Press. And Kymlicka argues that we must grant special rights to certain transitional or waning groups and to members of these groups in order to ensure equality among individuals’ contexts for choice. (See Note 16) But as John Tomasi has argued in response to Kymlicka’s view, stable cultures can also sharply constrain one’s context of choice by closing off a wide range of options that are outside of the particular culture. Furthermore, if we eliminate the condition of stability on what counts as a culture and simply maintain an existential interpretation of culture, Tomasi argues that it is not plausible to see anyone as lacking a culture. Even transitional or waning cultures are still cultures in this uninteresting sense, perhaps hybrids of a sort. On this interpretation of culture, it would not be clear how any cultures would be at a disadvantage. (“Kymlicka, Liberalism, and Respect for Cultural Minorities”, 1995, Ethics, 105, no. 3)

208

Both categories of collective rights would be ruled out by the counterfactual

test. This means that even if the case could be made that the right in question would

belong to individuals, the test would exclude such things as preserving a minority

language, because the latter sort of actions would need to refer to the interests of more

than one individual with the relevant identity. The test therefore does not simply

restate the normative individualism of both liberalism and rational choice theory, as

discussed in the Introduction. While the test does require the condition that the rights

and privileges in question belong to individuals, it requires more than this.32 The

view here is also stronger than the liberal presumption against harm; for it is not clear

that activities such as preserving a language would cause harm to anyone. Thus the

test does not merely reiterate existing liberal intuitions.

The test is however constrained by an important liberal principle. Earlier in

the chapter I said that any liberal account of respect for social identity must do more

than simply caution or assert that an action aimed at recognizing some identity-group

should not weaken the autonomy of the individuals in the group. It must be able to

specify a way to recognize social identity that does not itself undermine the liberal

sense of autonomy, and in particular basic civil and political rights. Otherwise, the 32 See Gutmann, chapter 1. It seems appropriate that the condition discussed here would be stronger than the condition that basic rights and liberties belong to individuals. First, whereas basic rights and liberties are universally applied, special identity rights and privileges would of course only apply to some people, and we would need to have assurance that identity recognition would not harm the autonomy of these people relative to others. More to the point, in order to best ensure that recognition for identity would not harm the liberal sense of autonomy, it seems important that such identity rights and benefits not only belong to individuals but also maximally allow for formulating and revising identity conceptions. Exposure to other customs, communities, and cultures is important to that end. Many rights and privileges that would belong to individuals but which would need to refer to the interests of more than one person, such as special language rights and exemptions from public education requirements, would tend to isolate cultures from one another, and so would not be maximally conducive to promoting the liberal sense of autonomy. The view offered here not only provides a means of safeguarding against the violation of basic rights and liberties, but is also responsive to claims about how cultures are not pure, unchanging things.

209

account could not be said to adequately show how it would enhance rather than

diminish autonomy overall—the expressed aim of such an account. This was the

liberal autonomy constraint.

I have demonstrated how the account here, with the distinctions in identity

form and correlative counterfactual test, specifies such a way. In particular, the

counterfactual test provides a practical procedure for ensuring that the liberal sense of

autonomy is not placed at peril. Because the test would exclude collective rights, we

would have assurances that an action would benefit particularized identity, even if it

also happens to simultaneously benefit the collective one in certain ways. Since an

action would have to be able to benefit an identity-based interest of any single

individual in the absence of others with the identity, it would not promote collective

interests at the expense of the basic liberties and rights of individuals with that

identity.

This point is another way of appealing to the fact of multiple identities. Under

the test, any action that would benefit a given identity would not infringe on the other

social identities individuals may have. A right to prayer-breaks would not

compromise a Muslim woman’s identity as a woman. The account here thus more

than satisfies the liberal autonomy constraint. It does not trade-off the liberal

conception of the nature of autonomy for other conceptions.

5.7 Conclusion

Because identity is so frequently, if not always, cashed out in terms of the

identity of groups, liberal political theorists tend to view respect for social identity as

210

giving status to groups as such, and frequently, as giving priority to groups over

individuals. My arguments have shown a way in which liberals can consistently

respect social identity. The prayer break example demonstrates that a social identity

can be actively respected without undermining the basic rights and liberties of

individual Muslims. It is an example of an action that would respect the Muslim

identity belonging to particular persons. However, an important theme of this chapter

has been to show that respecting social identity as part of the respecting of persons

would not support more robust communitarian positions endorsing special collective

rights and privileges. There may be other important ways of valuing identity, but for

the purpose of including it in the concept of a person and a person’s autonomy, we

must distinguish between a collective identity and particular individuals’ identities,

and include only the latter.

211

Bibl iography

Andreoni, J. 1990. Impure Altruism and Donations to Public Goods: A Theory of Warm-Glow Giving. Economic Journal 100: 464-77 Appiah, Anthony. 2005. The Ethics of Identity. Princeton: Princeton University Press. Aristotle. 1932, 1960. The Rhetoric of Aristotle. New York: Appleton Century Crofts. Axelrod R. 1984. The Evolution of Cooperation. Basic Books, New York. Batson, C.D. 1991. The Altruism Question: Toward a social-psychological answer. Hillsdale, NJ: Erlbaum. Batson, C.D. 1992. Experimental tests for the existence of altruism. Proceedings of the Biennial Meeting of the Philosophy of Science Association, 2, 69-78. Beauchamp, Tom L. and Childress, James F. 1994. Principles of Biomedical Ethics, 4th ed. (New York: Oxford University Press. Benhabib, Seyla, ed. 1996. Democracy and Difference: Contesting the Boundaries of the Political. Princeton: Princeton University Press. Binmore K. 1994. Playing Fair: Game Theory and the Social Contract I. MIT Press, Cambridge, MA. Brennan, G. and Lomasky, L. 1993. Democracy and Decision. Cambridge: Cambridge Univ. Press. Brennan, Geoff and Pettit, Phillip. 2004. The Economy of Esteem: An Essay on Civil and Political Society. Oxford: Oxford University Press. Broad, C.D. 1952. Ethics and the History of Philosophy. London: Routledge and Kegan Paul. Butler, Joseph. 1983. Five Sermons Preached at the Rolls Chapel and A Dissertation Upon the Nature of Virtue, ed., Stephen Darwall, Indianapolis: Hackett Publishing Company, Inc. Condon, W.S., & Ogston, W.C. 1966. “Sound film analysis of normal and pathological behavior patterns.” Journal Nervous Mental Disorders, 143, 338-347. Damasio AR. 1994. Descartes' Error: Emotion, Reason and the Human Brain. New York: Grosset/Putnam.

212

Damasio, Antonio. 1999. The Feeling of What Happens: Body and Emotion in the Making of Consciousness, Harcourt Brace & Company. Davis, Mark. H. 1985. “Perceptual and affective reverberation components”, in A.B. Goldstein and G.Y Michaels (Eds.), Empathy: Development, training, and consequences (pp. 62-108). Hillsdale, NJ: Erlbaum. Dennett, Daniel. 2003. Freedom Evolves, New York, NY: Viking Publishing. Dimberg, Ulf. 1982. “Facial reactions to facial expressions.” Psychophysiology, 19, 643-647. Dworkin, Ronald. 2001. “What is Sacred?” in Bioethics, ed., by John Harris. (New York: Oxford University Press. Eisenberg, Nancy. 1986. Altruistic Emotion, Cognition, and Behavior. Hillsdale, N.J.: Lawrence Erlbaum Associates. Eckel, C.C. and Grossman, P.J. 1996. Altruism in anonymous dictator games. Games and Economic Behavior, 16 (2), 181-191. Ekman, P., Levenson, R.W., & Friesen, W.V. 1983. “Autonomic nervous system activity distinguishes emotions”, Science, 221, 1208-1210. Fehr, Ernst and Gächter, Simon. 2002. Altruistic Punishment in Humans, NATURE 415, 10, 137-140. Fehr, Ernst and Fischbacher, Urs. 2003. The Nature of Human Altruism, NATURE 425, 23, 785-791. Fehr, Ernst and Fischbacher, Urs. 2004. Social norms and human cooperation, TRENDS in Cognitive Sciences Vol.8 No.4, 185-190. Fehr, Ernst. 2004. The Neural Basis of Altruistic Punishment Science 305, 27, 1254-1258. Feinberg, Joel. “Psychological Egoism.” In S. Cahn, P. Kitcher and G. Sher (eds.), Reason at Work, pp. 25-35. Sand Diego: Harcourt Brace and Jovanovich. Flanagan, Owen. 1991. Varieties of Moral Personality: Ethics and Psychological Realism. Cambridge, Mass.: Harvard University Press. Foot, Philippa. 1985. “Utilitarianism and the Virtues,” Mind 94. 196-209

213

Frank, Robert H. 1998. Passions Within Reason: The Strategic Role of the Emotions. New York: W. W. Norton and Company, Inc. Goodin, R. E. and Pettit, Philip (eds.) 1995. A Companion to Contemporary Political Philosophy. United Kingdom: Blackwell Publishers. Goodin, R. E. 1990. ‘Irreducibly social goods—Comment I’ in Rationality, Individualism and Public Policy, ed. G. Brennan and C. Walsh. Canberra: Australian National University. Green and Shapiro. 1994. Pathologies of Rational Choice Theory; a Critique of Applications in Political Science, Yale University Press: New Haven. Gutmann, Amy. 2003. Identity in Democracy. Princeton: Princeton University Press. Hatfield, Elaine and Rapson, Richard L. 2004. “Emotional Contagion”, in Larissa Z. Tiedens and Colin Wayne Leach (Eds.), The Social Life of Emotions (pp.129-43). Cambridge, UK: Cambridge University Press. Hausman, Daniel and McPherson, Michael. 1996. Economic Analysis and Moral Philosophy. Cambridge: Cambridge University Press.

Hayek, Friedrich von. 1942. “Scientism and the Study of Man I,” Economica, 9: 267-91

Hirschman, A.O. 1999. Alchemies of the Mind: Rationality and the Emotions. Cambridge: Cambridge University Press. Hirschman, A.O. 1977. The Passions and the Interests: The Political Arguments for Capitalism before its Triumph. University Park, PA: Pennsylvania State University Press. James, Williams. 1890. The Principles of Psychology. Henry Holt and Company. Kant, Immanuel. 1785/1983. Grounding for the Metaphysics of Morals. in I. Kant Ethical Philosophy Trans. James W. Ellington,. Indianapolis, IA: Hackett Publishing Co. Kymlicka, Will. 1989. Liberalism, Community, and Culture. Oxford: Clarendon Press. Kymlicka, Will. 1995. Multicultural Citizenship: A Liberal Theory of Minority Rights. Oxford: Clarendon Press. Lange, C.G. 1922. “The Emotions”, in C.G. Lange and W. James (Eds.), Baltimore: Williams and Wilkins.

214

Mill, John Stuart. 1859/1975. On Liberty. David Spitz, ed. New York: Norton. Monroe, Kristen Renwick. 1996. The Heart of Altruism: Perceptions of a Common Humanity. Princeton University Press. Nagel, Thomas. 1978. The Possibility of Altruism. Oxford: Oxford University Press Nozick, Robert. 1974. Anarchy, State, and Utopia, New York: Basic Books. Olson, Mancur. 1965. The Logic of Collective Action. Cambridge, MA: Harvard University Press. Plato. 1968. The Republic. New York: Basic Books. Prinz, Jesse. 2004. Gut Reactions: A Perceptual Theory of Emotion. New York: Oxford University Press. Railton, Peter. 1984. Alienation, Consequentialism, and the Demands of Morality,” Philosophy and Public Affairs, Vol.13, No.2, Spring. Rawls, John. 1971. A Theory of Justice. Revised edition (1999) Cambridge, MA: Harvard University Press. Rawls, John. 1993. Political Liberalism. New York: Columbia University Press. Raz, Joseph. 2001. Value, Respect, and Attachment. Cambridge: Cambridge University Press. Riker and Ordenshook. 1968. “A Theory of the Calculus of Voting”, American Political Science Review, 62, pp. 25-42. Rosenberg, Alexander. 1992. Economics: Mathematical Politics or Science of Diminishing Returns, University of Chicago Press. Rosenberg, Alexander. 1992. Economics: Mathematical Politics or Science of Diminishing Returns, University of Chicago Press Rosenberg, Alexander. 1998. “Folk Psychology” in J. B. Davis, ed., Handbook of Economic Methodology, Aldershot, Elgar; Rosenberg, Alexander. 1996. "Philosophy of Economics" in Audi, R., Cambridge Dictionary of Philosophy, Cambridge University Press.

215

Rousseau, J.J. 1964. "On the Origin and Foundations of Inequality Among Men". In R.D. Masters (Ed.), First and Second Discourses. New York: St. Martin's Press. Sandel, Michael J. 1982 Liberalism and the Limits of Justice. Cambridge: Cambridge University Press, 2nd edition. Schmidtz, David. 2001. “A Place for Cost-Benefit Analysis,” Philosophical Issues II, 148-71. Blackwell Publishers, Inc. Smith, Michael. 1994. The Moral Problem. Massachusetts: Blackwell Publishing. Gilligan, Carol. 1982. In A Different Voice: Psychological Theory and Women’s Development. Portland, OR: Book News, Inc. Smith, V. 2003. Experimental Methods in Economics. Encyclopedia of Cognitive Science, Lynn Nadel (ed.), New York: Nature Publishing Group, Macmillan Publishing, 1070-1079. Sober, Eliot and Wilson, David Sloan. 1998. Unto Others: The Evolution and Psychology of Unselfish Behavior, Cambridge, MA: Harvard University Press. Sterelny, Kim. 2003. Thought in a Hostile World: The Evolution of Human Cognition. Blackwell Publishing Stocker, Michael. 1976. “The Schizophrenia of Modern Ethical Theories,” Journal of Philosophy, 73, 453-66. Taylor, Charles. 1989. Sources of the Self: The Making of the Modern Identity. Cambridge, MA: Harvard University Press. Taylor, Charles. 1992. Multiculturalism and the Politics of Recognition. Princeton: Princeton University Press.

Weber, Max. 1968. Economy and Society, ed. Guenther Roth and Claus Wittich, Berkeley: University of California Press.

Williams, Bernard. 1973. “A Critique of Utilitarianism,” in J. J. C. Smart and Bernard Williams, Utilitarianism: For and Against. Cambridge: Cambridge University Press. Williams, Bernard. 1981. Moral Luck: Philosophical Papers, 1973-1980. Cambridge: Cambridge University Press. Wolf, Susan. 1982. “Moral Saints,” The Journal of Philosophy, Vol. LXXIX, No.8, August.

216

Young, Iris Marion. 1990. Justice and the Politics of Difference. Princeton: Princeton University Press.

217

Biography

Sahar Akhtar was born in London, England to Hanif and Zahida Akhtar. She

received her bachelors of arts from George Mason University, in economics. After

working for several years in the federal government, she went on to complete a

doctorate in economics, also from George Mason, in 2001. During 2007-2008, she has

been a Post-doctoral Fellow in the Political Theory Project at Brown University while

she finalized the present dissertation. In the fall of 2008, she will begin a position as

assistant professor in the department of philosophy at the University of Virginia.